Loading summary
Transcript56 lines
- [00:00]
A
If you have never heard of the tool Tailwind, let me explain it to you. They are a business that started all around Pinterest scheduling, Pinterest analytics, and they merged to much more than that. But they've gone back to their roots and they have done a four part study analyzing 1 million pins. And in this study they take 33 points of data and they analyze how you can grow on Pinterest, how you can go viral, which is such a great buzzword for people, but they're actually answering that question, how can you go viral on Pinterest? So I interviewed the CEO and co founder Danny Maloney about what he learned from this study and how it challenges our norms. I spent two hours, as you'll hear me say in the episode, just going deep into this study and there were some things that really significantly challenged my long held beliefs about Pinterest marketing. We get into that halfway through the episode. If you walk away from this episode and you think I want to learn more about Pinterest marketing or I want to hand off my Pinterest marketing to experts, you can learn more by clicking the link below in our description. Learning more about our services, you can go to simplepenmedia.com services. We offer everything from one time management to ongoing management to just consults to whatever it is that you need. Also, I have worked with Tailwind since 2015. We have used it as a part of our core strategy for our clients. It's a scheduling tool that we use, but it gives us a lot of great data. So I'm really excited that they have this study. Excited to talk to Danny. There's also links to the study below in the description, as well as the original interview I did with Danny about his origin story, which was part of our Summer Story series a couple of years back. This is a longer than normal episode. It is over an hour and the reason it is that long is because it took that long to discuss these parts of the study. There was a lot of things that were included in their study. Obviously it's 1 million pins and so I wanted to give it enough time to ask questions about why they did it, ask questions about their takeaways, and then summarize and talk about what what my takeaways were as well. So I hope you glean a lot from this episode and I hope that you'll also read both part one and part two of the study. Part three and part four will be out as soon as they can get the resources to distribute that part of the study. I'm sure it will be even better and go even deeper than what we already have available. Alrighty, let's dive into the interview. Foreign.
- [02:52]
B
You're listening to the Simple Pen podcast, Pinterest, for business advice that goes down smooth and easy. Here's your host, Kate All.
- [03:06]
A
Hey, Danny. Welcome to the Simple Pen Podcast.
- [03:09]
B
Thank you, Kate. Great to be here.
- [03:10]
A
I am so excited to talk to you about this study. And actually, I have to to admit, I spent about two hours in a coffee shop going through it, making notes, pulling up these things just to make sure that I was not only understanding where Tailwind was coming from, but the impact this had for those in the industry. And for those who are new to the Simple Pen podcast or new to you, new to Tailwind. Just so you know, Danny's origin story of how Tailwind came to be is something that we recorded a while ago. That will be in the description below for you. But I guess one of the things I want to start with is as I have been a tailwind user since 2015, I've watched this iteration of your company go from I knew you as a Pinterest scheduler, analytics, and then the addition of Instagram and Facebook and then kind of to this all encompassing tool. In fact, your tagline is now a tool that feels like a marketing team. And so for me, on the sidelines, it felt like a little bit of a move away from Pinterest and then kind of back into Pinterest, especially going deep with this particular study. Is this accurate? Am I summing it up the right way and why the move back to going deep with Pinterest?
- [04:24]
B
Yeah, first, I want to say I'm glad the research is proving valuable enough for you to spend that time on it because it's taken a lot of time on our end. So our, our hope is that it would be helpful. Yeah, I think there's a lot that's happened maybe over that five year period. If I go back to the beginning of it, there were kind of two things that happened simultaneously. One was we were hearing a lot from our customers that they were happy and having success on usually one marketing platform. Right. And that was good to get small businesses off the ground to build some traction. It was then a whole nother level of struggle to figure out how to go from one successful channel to. To multiple channels and then multiple successful channels. And so we were hearing that problem a lot. And so we were kind of reaching a point at the time where we were like, okay, how much more is there that we're hearing specific to Pinterest or Instagram that people want us to build and add to the platform versus this theme of trying to help people more broadly. And we were hearing a lot more of people wanting us to help more broadly. At about the same time, Pinterest began their Idea Pins experiment, which I think ended up having massive ramifications on the Pinterest ecosystem for about four years, until finally this past year, they started sunsetting Idea Pins and said, okay, we're moving away from this and back towards a more classical version of what content looks like on Pinterest. But during that period we saw really significant disruption in terms of creators and small businesses feeling like they knew what content to create for Pinterest. And then I think just another major theme during that period was people didn't like it because it was trying to keep people on Pinterest much more from an engagement perspective and sending less traffic out. And the primary reason most creators and small businesses were investing in Pinterest is the first place was to get traffic. And so there was this multi year period of conflict there that from our perspective actually led a lot of people to stop investing in Pinterest for a while. And it became this really big transition point. And I think now in hindsight, new CEO, new executive team at Pinterest, new chief Creative Officer, they've obviously made the choice to put that period behind them and move in a different direction. And I think that's probably best for everyone. But with that, we're now seeing a resurgence in interest in Pinterest because people are looking at it again saying, oh great, now that it's not all about Idea Pins anymore, we can actually get traffic here again and it's worth it for us to invest in creating content for this platform again. And so we're kind of seeing a flow of people back in, which in turn for tailwind has meant we're getting more feedback now about refreshing some of our functionality around Pinterest and adding to it for what's now possible today. So those are kind of, I think those are the two major themes going on. But for us, this study was really important to ask the question, okay, now that that shift in the content ecosystem has happened and so much time has passed, we probably need to reset and re benchmark what success is on Pinterest in order to make sure we're advising people properly and we're creating the right type of functionality that's going to help them actually succeed.
- [07:56]
A
Yeah, I like your summary of that too, because I do agree in this four to five year Period with Idea Pins, there was so many disgruntled people, so frustrated. And then you have at the end of that, this huge change with Google, with the introduction of AI and chat, GPT and Gemini and all of that, and it feels like this huge upset to where Pinterest can feel like a little bit of a safe harbor in going back. And I don't know about you, but I definitely felt like when I watched Pinterest Presents in October, I felt that feel of, we're going back to what we know and we understand the last couple of years have been difficult, but we're willing to move forward and push into that. So I can appreciate that tailwind at the same time was noticing that we were noticing that here at simple pin 2 as well, to kind of bring it together to kick it off, I would say with this study, which I guess leads me to. When you hear people talk about research scientists and doing studies and even I've wrestled with this in the past, you enter in with these biases that you go into the situation with. And a great example might be in the past. I know I've been frustrated with people saying live pending versus using a scheduler, as I'm sure you have too. And it's this thing of I'm setting out to prove that a scheduler is worth it. And I know I have the data, I know I have all this stuff. But truthfully, what's behind it is me trying to prove a point. Right. And so there's that delicate balance of when you do a study to. To find that, how did you guys really hold onto that at the beginning? Or was that a discussion to kind of challenge this longly held belief about Pinterest and how it works?
- [09:40]
B
Yeah, I mean, I think when you. Well, in part, I'm a data nerd, right. So, like, whenever I go into any type of data analysis or study, I think you have to approach it from a scientifically neutral perspective. Right. You just have to say, okay, I want to learn what the data has to tell me and question what the data has to tell me. If I go in with an agenda, I'm potentially getting the wrong answer out of the gate. Right. And that can be hard to do. But what was interesting for us in this moment is, as you may have seen, with some of our recent launches and a lot of what we are working on and have in the roadmap, we weren't really coming at this from the perspective of we're trying to justify the product that existed five years ago. Right. It's much more a question of, okay, if we're going to now reinvest significantly in Pinterest as a channel in refreshing the capabilities and expanding the capabilities, we want to make sure we're doing that in the direction that's actually going to create value for people, because otherwise we're wasting our time and energy by building the wrong thing. Right. And so it was actually really important for us also to go in, in an unbiased way and just say, like, hey, let the data tell us what it's going to tell us. And realizing that we are in a position of significant transformation, especially with the coming of AI and then from there we can make decisions about what to build. Right. As opposed to saying, we built this thing and we're trying to justify it with data, which is the wrong direction to go in.
- [11:11]
A
Right, Right. Yeah. So it has two parts for you. It has not only the discovery about Pinterest, but the discovery about what your tool can integrate in order to serve the broader community. Like you. You discover so many different things than just Pinterest data.
- [11:25]
B
And some of them are big findings where it's like, okay, there's this entire maybe area of functionality missing. Right. And we need to invest in that area of functionality that now is very important. And some of it might be much more niche, like when you start looking at what colors are generally used in content, and then we can come back and say, okay, do we need to relook at our create tool? And even things like which standard color palettes are we even making available to people in order to guide them in a direction of what works in the platform? Right. So the second one, you know, might. It might be a big deal from a feature building perspective. It's kind of a smaller tweak. Right. But then the, the former category there could be like, okay, here are entirely new categories of product that we now need to invest in in order to align with the future of where Pinterest is going as opposed to the past of where it was.
- [12:19]
A
Yes, a hundred percent. And I'm excited to get to that color thing because that I stuck on that for a little bit. Stuck. Not stuck. Stuck on that. Because it would just. It challenged a lot of my norms. So before we dive into the study, I want to say one thing to the listener. When you read these, read them with a fresh mindset. Read them. Challenging your norms. Read. And that's how I took it, was I want to read this with an understanding that this is 1 million pins analyzed. This has a lot of data and Even though I have 11 years of experience, I want to be new. I want to it, like, see this with a fresh perspective. And I feel like that it did that for me. So two questions before we dive in. How many hours did your staff spend on this and you?
- [13:06]
B
Yeah, I don't know the exact number. If I had to guess, we're probably ballpark, like, 150 to 200 hours total. So, you know, from designing the study to actually getting the data, extracting the data, cleaning the data, analyzing the data, and then figuring out how to summarize and write it up, it's got to be in that range. I think ultimately it is worth it as long as people are finding it helpful. Right. And we're definitely, at minimum, internally, we're finding it helpful in terms of informing some of those directions about what do we build next? What do we change next? What do we improve next? So there. There are immediate takeaways that we've got from the data there.
- [13:45]
A
Okay. Was there one thing in this study that completely threw you off?
- [13:50]
B
Ooh. So one of the really, like, maybe the biggest finding that has stuck in my mind is the data around concentration of engagement in a very small percentage of highly viral pins. Right. And I think part of that is that historically, my thinking around Pinterest was more like, you're building a portfolio of pins over time, and sure, there are going to be some outliers, but as long as your average pin keeps getting stronger over time, you're heading in the right direction. Seeing the degree of concentration that exists now, and that is promoted by the. I shouldn't say the algorithm, but the net effect of all algorithms on the platform. Engineering pins, maybe that's the wrong phrase, but, like, designing content right in order to have a better chance of being among those viral few is probably where a significant amount of ROI lies. And people, I think, intuitively have seen this in their analytics in the past when they have, like, that one superstar pin that they just can't ever seem to replicate, but it's such a big win that it drives tons and tons of traffic every month. Getting those superstar pins is probably the key to massive amounts of roi. You can still have meaningful ROI without them. But, yeah, that data really stuck with me and stuck with me in thinking about what functionality do we need to add to Tailwind or offer that will help point people in that direction more aggressively so that they can really satisfy the content need that exists on the platform?
- [15:27]
A
There's so much I want to go into with that. Okay, we'll get to that. But I love that that's what you found because I think I always find that interesting too. And a common phrase that I hear from people when I talk to them is I looked in my analytics and the same pin is popping up over and over and over again. And there's almost this frustration with instead of celebrating what it is and how you can create spinoffs up from it and all of these things. So it is fascinating to me that it keeps going.
- [15:53]
B
And most likely that superstar pin is probably one of the top ranking pins on a significant search volume search. Right. So I would celebrate it the way I would celebrate getting a top three ranked article on Google for a significant keyword search. Right? Yeah. And you don't want to lose it. You want to retain that as long as you possibly can and keep getting that traffic. It's hard to get a top three ranked article repeatedly on a given search though. And it's hard to get a top three ranked pin repeatedly on a given search. Right. So yeah, I'd say be grateful for that, that superstar former and then keep trying to add to it and maybe you can create others alongside it that just elevate traffic even more.
- [16:39]
A
Yeah, great words of wisdom there. Okay. There are four parts to this study. There are two that are published at the time of recording and I want to talk about part one because this is the success for organic marketing strategy. So we're just talking about non paid. You said your four key findings are Pinterest is still an evergreen content platform that has not changed when viral pins peak. Fresh pins, you know, I love that word. We're going to talk about that drive most new traffic from Pinterest. That's interesting. And winning viral pins are the key to success on Pinterest. So let's take this in the four parts evergreen content that doesn't feel like a big earth shattering surprise. But I like that we have more data around it. I'm seeing it from Pinterest, seeing it from you guys. And one of them was 60% of pins are over a year old. And I actually just read on Pinterest that their shelf life is 13 months. So with this evergreen, was that just a support for you guys to say yes, it's still, still is still search and discovery.
- [17:37]
B
Yeah. I think it's really important for people to fundamentally understand and one of the reasons why this question was important is because of the emphasis on fresh pins that started really around the same period as the idea pin period. And there was a valid question there of how much has the Algorithm shifted to value freshness. And then you get into like, what's coming in part three of the study, which is how is freshness actually defined and what does the data show about that? But I think this data actually does show a bias towards fresher content than would have been in the case. The case in the past. And by that I mean, if we did this same study five years ago, we might have seen a similar percentage of engagement coming from pins that were maybe over two years old or three years old. Right. So there is more of a bias, it seems, towards fresh content, but it is still fundamentally evergreen. Right. And where I think that is critically important, especially for people who are newer to Pinterest as a channel is going in with realistic expectations about how long it's going to take to see success. Right. In the SEO world for organic search, people now have accepted it takes generally six months or more to rank highly for competitive keywords, and that's if you even succeed in that effort. On Pinterest, we don't see as much of a broad understanding of that. And so we do tend to see more people who come in and they think month one they're going to knock it out of the park and get some viral pins and have their traffic going crazy. And in a lot of cases, that probably leads people to give up on the platform far too early. And then the challenge becomes if they give up on it and they try to come back later, they're kind of starting that cycle all over again fresh. Right. And so that break period can end up costing you a lot of time in terms of how long it takes to get to real traction. So that's why I think this data point is so critical and part of where we're starting to go with this data and a part of what I've been thinking about in this study is how do we help people benchmark newer pins versus more aged out pins in terms of performance? Because if you compare brand new pins to aged out pins, it's an unfair comparison. It's apples to oranges. Right? What you really want to understand is early in the life of a pin, are you getting signals that these pins are likely to be as successful or more successful than your earlier generations of pins, and if they are, keep going in the direction you've been going creatively and strategically, and if they're not, try a new path. Right. So there's something to the analytics here that I think needs to evolve into more of a cohort or generational type of view of analytics, which I don't think has really existed anywhere, including within Tailwind, including Pinterest's own native analytics in the past.
- [20:36]
A
Yeah. And I think there's only that setting where you can see anything that's happened within the last 30 days. But then I think then you have this gap of 30 days and six years. Right. So it's like there's that middle section and you're like, well, which goes to actually the second part of your study of when viral pins peak. You had said viral pins peak between one to two years. So this analytical gap ignores that, especially if you are somebody who has committed to Pinterest. You're saying, I'm going to give it a full year, maybe even two years, because that's when I can see what is actually working. So you had some findings with your oldest pin in your data set, which was from like 2010. So, I mean, I think we're kind of hitting the point home here that, like, your older pins do great work. Your new pins can do great work, too, but you need that middle time frame to really understand how they're working.
- [21:32]
B
Yeah. And really, it's almost a continuity of life cycle. Right. Of the content is the type of view we probably need to get to to make the data really useful. But that's a. It's a challenge to get to because those pins from the last 30 days, first, you want to know what to benchmark them against. Right. But you also want to look at days 31 to 60 and 61 to 90 and 91 to 120 and so forth. And really across that to understand if a cohort of pins is really performing or not performing, you kind of have to look at that type of a timeline view to see how it progresses over time. Not just doing last 30 days and everything. Right, right. That's. It's the everything bucket is far too broad.
- [22:17]
A
Right. It is, for sure. So I'm, I'm great. I'm grateful that you all are figuring out this gap and hopefully we'll solve for something, that analytical gap. And I think since we're already on this discussion of, you know, you brought up the word fresh pins, which came around in 2020. We've been asked about it a lot. We did some research into that first interview you did with Tailwind or with Pinterest, where they brought that up. And it was this whole wave, I think, of that point of there became this trend of people saying, okay, I've been on Pinterest for a while. I'm so exhausted. I'm just going to Create new images for old content. And what we saw with a lot of our clients was that when they did that, they decreased in performance over time. But those who had a, you know, update the old but build the new had an increase over time. But it was hard to convince people of the fact that they couldn't just go on this plug and play. I'm going to redo 25 different images every day to where the. The pin or that phrase overall was kind of like this phrase drives me crazy. But I'll. You brought it in here to the study. Fresh pins drive most of the traffic. So you give me your new perspective. 2025, the distinction, fresh pins. And then we also. I read the word creates in here too. Creates versus saves. This is a Pinterest term. So let's talk about this. Where are we at?
- [23:41]
B
Yeah, I think the real distinction that Pinterest made fundamentally was creates versus saves, right? Creates being pins that you've created and pinned for content pointing at your domain that you have verified and that you own versus saves, which previously were called repins, but basically a pin of someone else's content or a repin of someone else's content, if you want to think of it that way. They created this dynamic a few years back where creates and saves are fundamentally different. And one of the goals of that was to funnel distribution towards creates, right? Which I think is ultimately a very good thing because it helped solve certain problems people had in the past. Like I'm the creator of this pin, someone else repinned it or saved it and their copy of it is getting all the traffic. So they're getting a bunch of followers and a bunch of engagement and I'm not getting any on my profile. Right. So this distinction was important to help creators in driving more of the attention to their brand, their profile, so they could build an audience over time. The fresh pins dynamic, I think was often used as a substitute for creates and kind of blended the two. I think the world we're living in now and what the data will show, especially when we get into the part three of the study, is that within Create there are degrees of freshness. Right. So the freshest version of a create might be a brand new URL, brand new page that you're creating a pin for for the first time. Right. A less fresh than that, but still pretty fresh, is an existing URL or page that Pinterest knows about, but you're creating a new image, a new pin to frame that content for a different audience, different keyword set, whatever it might be less fresh than that is going to be. Same URL, same image, new description. Right. And less fresh from that is going to be all the same, but a new board or something like that. Right. So I think that's where what we see in the data that'll be in part three is that as you go down those degrees of freshness, the decay in performance happens a lot faster and so you want to stay as close as you can towards higher levels of freshness. But that said, if you talk to any content creator or website owner, the idea that if every single PIN had to be a brand new URL on their website with sufficiently unique and original content on it, that's way too much investment in a single PIN based on the ROI you're going to see on that pin. So you kind of have to blend in some of the new images to an existing page, maybe sometimes using the same image but different text elements on the pin in order to help Pinterest understand it in different lights. So you probably need a blend of those different elements in your strategy. But a lot of that is driven more by ROI on your time than it is necessarily on optimal performance. Because really you're not just trying to optimize performance, you're trying to optimize performance relative to the time and money you spend to get it right.
- [26:54]
A
And ultimately that connection. Right, with the pinner. Yeah, sorry, I'll sort interrupt you.
- [27:00]
B
Oh no, absolutely. But yeah, I think that's a really interesting part of the analysis. I do think with AI we're seeing some interesting new type of disruption there where like there are some people in the ecosystem who are basically promoting creating AI driven content farms. And in the near term that might be an arbitrage worthy strategy. I believe what we will see, as has happened with past cycles of content farms, is eventually the platforms will figure out how to snuff them out. And so maybe they'll make a little bit of money for a while and then eventually they'll just go to zero and go dark overnight. So I would warn people to not go too far to the extreme there of interpreting this as saying, oh, success equals generating as many pages as possible on my website, even if they're bad pages. Right, right. That is a bad idea. That that will never be a sustainable return. People will who work at not just Pinterest, but all of the major platforms out there, they will figure out how to weed out those sites and then all that work you put into building them basically dies off overnight. Right. So there's gotta be a balance here of making sure it's quality traffic, not just trying to play the algorithm, so to speak.
- [28:17]
A
Well, my last podcast was about that AI and Pinterest images, because I saw this really coming, you know, into the fold and have the same assessment. It's kind of what I see it as, is there are these people who are looking to build long term sustainable people first businesses instead of a money first business. And I. You've seen it on Google, you've seen it everywhere. There's always these loopholes that can be exploited to go after money quickly, right? And like you just said, they figure out how to close those loopholes because they want to generate helpful content for their users, because they want their users to stay on the platform similar to what they saw with idea pins, that it was this great disruptor to the natural ecosystem of the platform. AI images and crappy content can have the same effect over time. So it's in their best interest to try to get rid of it. But I think it supports what you've said here that there's this combination of both the viral pin and the new content. And I think the same thing of what we've seen, and this was how you summarized part one, like winning viral pins are the key to success. And I took this phrase, you said top the top 1% most viral pins accounted for over 50% of total impressions and clicks across all fresh pin pins published in the last 90 days. And you said, and here we have a number. And this is yet another reason why Pinterest recommends that creators try to publish as much as 5 to 25 new fresh pins per day. Given what you've just said, that's that healthy mix of kind of repurposed and new. Am I correct in that assumption?
- [29:54]
B
Yeah, that's exactly right. And I think there's a couple of things that jumped out at me. So. So that 1% driving 50% of the total impressions and clicks, that was that one finding at the top that really stood out to me. And that is specifically within fresh pins within the last 90 days. But thinking about the idea of analytics moving to more of a cohort or generation style of looking at the data, another way of putting that is there's a bunch of content posted in a period. Pinterest is going to test all of it. A lot of it's going to be weeded out very quickly because it doesn't generate engagement and that content's going to die quickly. Right? Some of it will get a larger audience, some of that will get a larger audience, some of that will get a larger audience. But I think what this data point shows is that those breakout pins that earn the larger audience through multiple cycles of that algorithm can elevate to really substantial audiences pretty quickly. The flip side of that is the majority of pins will generate little to no engagement over their lifetime. Right, Right. So I think that's where the 5 to 25 idea comes in, is if you're creating too few pins, variance in the, in the results is going to hurt you too much and it's going to make it too unlikely that you find success. So you want to be creating enough content to, you know, for lack of a better phrase, buy enough lottery tickets in that lottery. Right. Where you're going to get some winners and some might be, you know, $20 winners and some might be jackpot winners. Right. But you want to create enough content to give yourself enough chances to have some pins that surface and then keep doing that over time. But if you're doing it purely for the sake of volume and it's bad content that's not engaging or it's 100% AI generated disingenuous content, in the long term, that's not going to be a winning strategy because that, you know, that will eventually come to an end. Right, Right. So, yeah, there's that balance there. I think for each business, it becomes a question of, based on the specifics of their business, how much content, how much quality content can they realistically generate? And that becomes the right number. Right. So you want to maximize that number, but you don't want to artificially push past it and start creating lots of really low quality content.
- [32:17]
A
Right. Like more is not necessarily better here than, but connecting. And I think that's an important thing to take away. Like we're talking about data and numbers and how to go viral. The important thing is getting back to who you are talking to and how are you connecting with them. And I think that leads us into part two, which, you know, we, you started out talking with niches and my brain goes towards the Pinterest predicts and their, you know, what they put out every year, kind of marrying some of those two together with like, what are people already searching, what is your niche? What category are you in? Excuse me? And how can you kind of bring those all together? So you said this. There's a definite concentration in today's very top categories. The top six account for 80% of the most viral pins in your data set. And we're seeing a significant shift in the mix of content that is succeeding. So the Top six, Can you rattle those off your brain? I know.
- [33:16]
B
Oh, gosh, I'm gonna cheat.
- [33:19]
A
Look, cheat for sure. Yes.
- [33:21]
B
Fashion, decor, food, beauty, DIY, crafts, and art were the top six.
- [33:27]
A
Okay. And you said that food actually seems in particular to have taken it. Taken a step back from its prior dominant position as number one and maybe travels a little less prominent than you have historically seen. The drop in food seems a little surprised. Surprising to me there, but I guess that's kind of taking us down a little bit of a rabbit trail of food. But I would say, as far as how things go viral, would you say to somebody listening who's not in these top six or seven, is there still an opportunity for those within the outer ring to do well even though they're not in this top six?
- [34:05]
B
Yeah, I mean, there definitely is. And you know, the difference is going to be there's going to be less aggregate traffic in some of the categories that are further down, most likely. Right. That doesn't mean there's not still a lot of traffic. Right. So there's still a lot of traffic. Generating top pins in those other categories is still going to produce a great return for your business, a lot of visibility for your content, a big opportunity to connect with audiences. It's just that naturally, the bigger categories will have even more of that. They also might have more competition over time, though, because as people learn that those are the bigger categories, more people are going to try to create content for them. The food part surprised me also, and that's because we used to look at data like this pretty regularly early on in the life of the company. And there was a period of time where looking at this type of data point just became so boring because it was like every time and 60, 70% were food pins. Right. Of the top, most viral pins. And there was just so much traffic and activity specifically around food that it made everything else seem smaller. When I look at this data, though, part of what I see is Pinterest delivering on what they've been talking about publicly, which is the move towards shopping and commerce as a major driver of the platform. So I look at categories like fashion, decor, beauty, DIY and crafts that have risen in relative popularity. Those are also naturally more shoppable categories than food. And so some of the features that Pinterest has developed, things like shoppable pins and product recommendations. Right. Those advantages of those features in terms of engagement are going to accrue to pins in those categories, but less so to a category like food. And so I think within the algorithm that actually just creates a natural situation where these categories are rising because they're getting more engagement from the audience. And categories that don't benefit from those features are naturally taking a bit of a step back. That said, even if food is the number three category now, that's still a ton of traffic for food related content. Right. So it's still huge, but it's not the overwhelmingly dominant number one that we used to see, you know, year in and year out for so many years.
- [36:35]
A
Mm. So I want to. I. As you're talking and we're working through this with this part two of how to go viral, I'm going to take it kind of in steps as someone listening. Step one is really looking at where your niche falls and kind of, I would say, choosing to figure out, like, are you in the top six, Are you in the other remaining ones? And how do you want to capitalize for the viralness of your niche? I would say, like, kind of thinking about using Pinterest predicts or just really tapping into what are they searching. That's kind of the first thing that I take away from this. But then number two, we go into PIN formats. Right. And which PIN format to use. Because that comes to a second question of how do I get engagement? And a lot of people have had this video works. Video doesn't work. Like you said in the beginning, we've moved away from the idea pin format and we see this, like, simplified PIN format where it's just a pin is a pin, but we still see this image pin be the thing that is number one, and that's what your study kind of pointed out, was that image pins remain the top. That's what people click on.
- [37:44]
B
Yeah. I mean, that is Right. They also probably remain the top by volume also. So, like, it is also a little bit of a biased comparison where there are far more image pins out there. I think what's important about that data point is compared to what we saw during the idea pin period, where there were so few idea pins and they were having so much of the attention and engagement funneled into them, that was very clearly unnatural in the analytics. Right. And it was a clearly intentional decision to reserve distribution for idea pins in order to try to incentivize people to create more idea pins. But there was still that missing factor of if it's not driving traffic back to my brand, why. Right. But in this case, I think what's important is we're not seeing any clear bias towards other types of content. And so I think it's very Much true what Pinterest is saying a pin is a pin now, and create the right type of pin, the right type of content to communicate to the audience you're trying to communicate to, and to get a point across from an ROI perspective for most businesses and creators, I think that will lead to image pins. Just because image pins are so much faster, cheaper, easier to design than videos are to make, with the exception maybe of brands that already have very good, very engaging video content. No reason to not use it on Pinterest. It absolutely can work. It's just that if you don't already have it, it can be very expensive to figure out how to make it. When analyzing those viral video pins, one thing that we saw a lot of was pins, where through watching the video, you were able to demonstrate use of a relatively complex product or service, or you were able to show the process of creating something from end to end, demonstrating, you know, in the fitness space, demonstrating workouts and, you know, things of that nature, where if people see it, they're going to better understand how to do it in their real life as opposed to reading about it. Right. One example I love is like there was one account in particular that had a lot of content around building, like in home, I forget what the term they use were, but basically like in home, jungle gyms and these really amazing play spaces for kids. And it was one of those things where you, you watch the video and for me, this is something I never would have even thought of as being possible to put in a house. Right, right. But then you see the kids playing on it and how much fun they're having, and you're instantly like, oh, that's kind of cool. Maybe that's something that we should think about, you know, if it's appropriate for the kids and if you add the space. Right. So, you know, products and examples like that are where we saw video performing really well. But, but image pins very clearly were dominating the format again.
- [40:45]
A
So then the follow up would be text or no text. Raw images versus designed pins is what you labeled it in the study.
- [40:53]
B
Yeah. And what was interesting there was it varied a lot by category, so different norms within the categories. So, so food as an example, you know, another maybe obvious one was like the inspirational quote type pins. Right. Text overlays. Right. Pretty much everywhere, almost exclusively on the better performing pins. And I think that's because when people are looking at the image of a completed recipe, let's say they might think, oh, that looks good, they might not fully understand what it is that they're looking at and the text gives them a really quick, easy way to understand it. On the other end of the spectrum, categories like decor and art and tattoos. Photography itself as a category tended to be just the images. No logos, no text overlay at all. Fashion and beauty actually skewed in that direction. Also, the exceptions in those categories were around roundup posts. So 17 fall nail styles would tend to have text overlay and a designed image, as opposed to just showing one nail style because you can't instantly tell what the piece of content is about without the text. So we have a chart in the report where it breaks out category by category where they fell on the spectrum. But that might be something for people to look at if you're in a given category or even a combination of niches depending on your business, thinking about how you tailor the content for each niche to what people seem to be responding to within that niche.
- [42:30]
A
Yeah, agreed there. Okay, this, this was probably the most fascinating part and maybe more most challenging part of this part two in how pins go viral because this is color on pins. And particularly I think this challenged my idea of always having branded pins. Pins, right. So I have often wondered if our branded colors have really hindered us from growth. Also, I'm a B2B which we know is different than B2C and T's B2B. Like you get different levels of traffic and virality within even that alone. But this said it said bold colors of Your study indicates 87 of the top 100 hex codes identified as dominant colors in PIN images were white, near white or some shade of light gray. And essentially only three were bolder, easily discernible colors. And your call to action in this section was seeing this was interest or this was your call to. I think I wrote this right. Essentially I'll summarize it instead of trying to read for my notes. Don't always go with your branded colors because those colors might not be engaging and that your pin is a standalone and it's not necessarily within the silo of people knowing you, which we know that. I feel like I've said that a thousand times, but I think I've always connected it as I want people to see my standalone pin, see the branding and connect with it and then connect with the connection on my site like you're in the right place, if you will, because you've seen these kind of branding. But I think what you're saying in the study is maybe try some different colors that aren't within your branding in order to go along with, okay, if we want a viral pen, if we want a trending pen, then use the trending colors. That challenges a lot of my core beliefs about images.
- [44:29]
B
Yeah, that was kind of a gut punch for me looking at the data. So to your question of remaining unbiased in looking at the data, this was a surprise. I think there's maybe two different elements of this color part of the data. One is so much of the white light colors, light grays, et cetera. Some of that is actually going to come through purely through the RAW photography that is contained within pins, using white space, using negative space to frame objects of interest in the image, which is something that happens naturally in some cases, but also can be a very intentional choice to make something pop. So it probably is a little bit biased because of that phenomenon. But then there's also the piece about when you do design images instead of using RAW photography, and you do have the text overlays and elements on the images and templates that you're using and so forth. That was where I was really surprised to see that amongst the most viral pins, brand colors were actually incredibly rarely used. It was like 5% of the most viral pins. And the number one was black and white for, like, text overlay and design. Right. People just using black and white. Right. Which I guess you're scrolling quickly through content. It stands out, it's easy to read, there's high contrast. Right. So you can consume the information quickly. A very close number two, though, was using colors that match the colors in the photos. And that makes sense, I think, especially for seasonal content, where now you're thinking about the mentality of the user and it's like, okay, yeah, if I'm looking for fall recipe ideas, I'm in that frame of mind of seasonality. And so using colors that fit the season and fit the content likely is just going to make more sense to me as I'm looking at different content and deciding which content to actually engage with. So that part was interesting. This is one of those places where I look at things like our create tool within Tailwind, because we do have features that let people store their brand colors. And interestingly, in our historical research, that was one of the most requested features that people believe they should be using their brand colors. I think to some extent this data suggests maybe it's the opposite, maybe it's more important. Thankfully, we also have a feature that says match the colors in the photo.
- [46:58]
A
Right.
- [46:59]
B
Um, that feature is there already to be able to do it quickly, but maybe that and then providing people with some of the more successful color palettes that they could consider trying on for size for their brand. I think fundamentally what a lot of that comes down to, though, is there's a very big difference between our brands and Coca Cola or Apple. Right. Like, those brands are so globally known with so much marketing and promotion behind them, that maybe a very large brand like that actually gets more benefit from including its branding in a pan because it is so instantly recognizable and it provides that sense of authority. Whereas for smaller companies, it really is about getting people to buy into the content in order to then get them to your brand to then learn about your brand and over time, build authority and build credibility for your brand. Right. If you lead too heavily with the brand, ironically, what this data might say is you might put people off from ever coming to your site and engaging with your content, and then you kind of miss out on the opportunity to actually build rapport with your brand.
- [48:13]
A
Right. Okay. So the few takeaways I have, and even for myself, well, there's a lot. But you had a phrase in there. A pinner searching for fall decor ideas is in a fall mood, not a quote, unquote, your brand mood. So I think one of the things that you can challenge here for the listener to take away as far as the images. And even for me, I'll. I'll share what I would do right now is this is a tactical move on Pinterest to kind of. It's not a strategy change. It's just a tactic move to say, okay, for me, I use red. So maybe I move away from using something softer and challenging my norms and using different images for the next three, four, even six months in order to collect data, to say, I was doing XYZ for years and years and years now. I've tried this for six months. Here's what I've learned about these pins. And I'm still creating content. I'm still in the same flow like nothing's changed besides that factor. That would be a fascinating test for people to do. In fact, I think we'll probably do that just because I'm curious about it. But one of the things I don't want to take away is I don't want to remove my logo from it just simply because I want to keep some element of branding on there. So I think there is a tasteful way to keep that on there, but move with the colors and kind of make it kind of a cohesive look.
- [49:35]
B
Yeah. And this is not very specifically. This is not analyzing logo. No logo. This is Analyzed dominant colors in the design. Right. So you could still have the dominant colors in the design, be more appropriate to the image or the platform generally, and then still incorporate your branding in a way. Right. Yep. It. But what the, what this data does suggest to me is going too far in emphasizing brand in the dominant colors in the design, probably for a lot of small businesses is a mistake.
- [50:06]
A
Yeah, I, I agree with you. I agree with this study. I'm good at. Well, I don't agree yet because I don't have the data, but I will say you've definitely got my wheels turning. Okay. This kind of last element of how to go viral gets into the keyword piece, the search and discovery, specifically Pinterest descriptions. And one thing in that part that I'd like to call out was the most viral pins tended to have shorter descriptions. We've heard over the years that you can use up to 500 characters on your Pinterest descriptions. You need to max it out, not keyword stuff. But what you are finding is that it was potentially just one or two sentences, is that correct?
- [50:45]
B
Yeah. So the specific length, like the top 1%, most viral pins, just for comparison, top 1% were I think, around 220 characters on average, and the bottom 80%. So pins that really didn't generate much traction or virality at all were closer to 400 characters. Right. I think a little bit shy of 400. And basically when you look at the chart, it just descends gradually as you go higher in tiers of virality. So what we're not suggesting here is like a one word, six character pin description. But it was interesting to see this trend. My hypothesis on this might be too much, actually could confuse the categorization of the content and could dilute the key signals that you're trying to send in terms of how to think about and categorize this content, which might then make it harder for it to find the right audience. Right. So trying to rank for too many things on one PIN probably is the culprit here. And this is where you get into the degrees of freshness. And maybe it's better to design a different alternate version of the PIN for that page that is trying to rank on something else. Right. And that's where creating multiple versions of a PIN can be a virtue. Because now each individual version of the pin, which all should be relevant to the content they're linking to, ultimately may speak better to someone with a slightly different tweaked search or, or category that they're browsing. So I think that's probably why this type of trend would exist. I don't think we have enough data here to say that with certainty. But. But it's also, you know, it's worth calling out that shorter doesn't mean zero is good. Right? Isn't it? Yeah, it's just not. Not necessarily always feeling like you have to maximize the character count.
- [52:52]
A
Okay. You added in hashtags here, which listeners will know I've died on this mountain. But your study gives a more nuanced approach to them. So tell me a little bit more.
- [53:00]
B
So. So let me ask, which. Which side of the mountain did you die on?
- [53:03]
A
Oh, no. Hashtags. Get rid of them. Don't even talk about them. Like, I, I think it's just because Pinterest has indicated this. You know, we, we love them, we're going to use them, and now we're not going to love them, and then they wind up somewhere. And I think just recently I'd see somebody who said they loved using them, and I just think they're a distraction to somebody who's learning to be a good Pinterest marketer to say, like, okay, I'm going to write a really great description, but then to get distracted and so just the users don't use it. So it's been this thing of when I see other teachers teaching that it's a strategy, maybe I feel a little protective of the people who are receiving that strategy and incorporating the time into it, and I think it's just not worth your time.
- [53:46]
B
Yeah, and I think that's a, that's a fair argument. What we saw in the data was, I think it was about 1 in 4, five of the most viral pins included at least one hashtag. What we don't know is the denominator there. So out of all of the pins that these were competing against, what percentage of them included one or more hashtags? If I had to bet money, I would bet it's probably lower than that one in five that included any hashtags at all. But I don't know that for sure and can't say that for sure. What I think it does indicate, though, is that hashtags probably aren't penalized. Right. So it's, it's probably not a net negative to use hashtags. It does seem like. So what was interesting is a lot of the people who were using hashtags were using hashtags as their entire description or even like a majority of their description.
- [54:35]
A
Interesting.
- [54:35]
B
So maybe they would have like one line, but then they would have seven hashtags after it or something like that. And so my guess here is what's going on is Pinterest is reading the hashtags, probably removing the actual, you know, hash, Right. But then interpreting that as keywords. And so I don't know if that's any different than if the words were just used without the hashtag. But it does seem like it's probably no worse. Right. So it does seem like Pinterest can see through hashtags to indicate relevant keywords for content. It's hard to say if it's, if it's a net positive or net negative with the amount of data we have, but it doesn't seem to be as strong net negative.
- [55:20]
A
Well, and actually, now that I'm putting two and two together with Pinterest having an auto posting from Instagram feature set up for, for those who don't have their own website, it would make sense that they would have to come up with some kind of functionality for the hashtag, since a lot of those are like auto pulled with how people use them. So that really came up for me when you said seeing these with seven or eight hashtags, sometimes it's a repurpose and people are like, just throw it up there, right, and see how. But I appreciate that we're seeing this overriding of hashtags potentially into keywords. Now we don't know that for sure, but it, you know, it's not going to hinder you, not going to be a net negative.
- [55:57]
B
And part of, part of what drives that belief for me is that hashtags are ultimately searchable, right? So like, if you click on a hashtag, it will lead you to a search page for a relevant term. And so that's what leads me to think, okay, these hashtags have been converted into keywords somewhere on the back end in processing this piece of content and mapping it to the graph. And so, yeah, it seems like they, at some point they must have developed a process for understanding hashtags better. And makes perfect sense that the integration with Instagram would be an impetus for that.
- [56:32]
A
Okay, I've been going for an hour here and there's so many things that we've covered in just part one and part two. And I would highly encourage, encourage anybody to go read this and as you read it, take away just one or two things that you're going to change about your Pinterest strategy, or if you're totally new to Pinterest, taking it and creating one to two things you will do that will be Your non negotiables. Tell me about part three and part four and when we could expect them to be released.
- [57:01]
B
As soon as we can finish writing them up is the answer to the second question. Uh, this second part was actually by far the longest and meatiest, so that one took a lot of time. Um, so these will be a little bit quicker to get out. Um, part three is specifically focusing on the nature of fresh pins and so starting to help people understand for newer, fresher content. How do I think about benchmarking it? How do I think about what success looks like or doesn't look like? And I, I think that's really important for someone to accurately evaluate how they're doing on an evergreen platform like this. And then the fourth part is looking at more of account level norms, so specifically zeroing in on the accounts that did repeatedly create more viral pins on average, and trying to figure out, you know, what were their norms by different standards, to see if there's any guidance there of what a successful account might look like. But those are the two remaining parts. I think after that, we'll put this data to rest and then start planning for when do we refresh it again? How much time will be enough time pass that it's actually interesting to look at. Maybe it's an annual thing, I don't know, but it's been a lot of fun diving into, and I just hope folks are getting a lot of value out of it.
- [58:23]
A
Well, I thank you for your time and I know that it has been valuable to me and to my team. I think we are asking a lot of the same questions that you all are with this past four years of what Pinterest has been doing. What's the new Pinterest? What does it look like? How do we use it as marketers? How do we leverage it for our business? And studies like these give guidance that is data grounded instead of just anecdotally kind of grounded, which we appreciate because we want to go with that. We want to follow those pathways instead of rabbit trails that waste our time. And I hope that the listener will dive into it and it's enjoy that too. So thank you so much. I really appreciate your time.
- [59:00]
B
Likewise. Thank you, Kate.