← Marketecture: Get Smart. Fast.
Loading summary
Transcript61 lines
- [00:00]
A
Foreign welcome to marketecture where you can get smart fast with in depth interviews of leading technology executives. I am Ari Paparo. I'm joined today by Gaurav Chindler who is the founder and CEO of Tercept. Probably a company a lot of people haven't heard of, so we want to learn a bit about it. Gwarav, thank you for being here.
- [00:26]
B
Thank you so much Ari. It's a pleasure being here. Thank you so much.
- [00:29]
A
So quickly, what is tircept?
- [00:30]
B
Tircept is data infrastructure provider for ad tech and media companies. We've seen that far too many really good ad tech media companies struggle with substandard data infrastructure and we believe Tirsip can do a good job supporting them with building out their data infrastructure and bi.
- [00:48]
A
Oh yeah, I mean I've experienced that building a dsp, the data pipeline tends to be more work than the data itself. So let's start by telling me about the company. Where is it based, how many people, how long you've been doing it?
- [00:59]
B
So we are headquartered in the US we are Delaware C car we have a couple of folks in the US with sales roles and we have about 30 people in India who are in development, customer support, data analytics and stuff like that.
- [01:13]
A
Great. And so how did this come about? How long have you been doing it?
- [01:17]
B
Well, I've been in the itch space for over 15 years now.
- [01:20]
A
I'm sorry.
- [01:22]
B
Doorship is about 6 years old, but I've been in the itch space over 15 years. And before Doorsept I was running another company called Wizuri and Wizuri was one of the earliest companies. We started that with a couple of co founders in 2007, raised about $30 million of funding, built a lot of retargeting buy side use cases, dsp, dnp, that sort of thing. Interestingly, we were the first company in Asia at Wizori to integrate with Google Ad Exchange. We were the first company in Asia to integrate with the erstwhile Facebook Exchange. You would remember this.
- [01:53]
A
Fbx.
- [01:54]
B
Yeah, fbx, exactly right. So we were building a lot of buy side use cases, bidding recommendation algorithms, stuff like that. We did that for about nine years. In 2017, me and my co founders sold that to another company called Apple, which is a listed ad network based out of Singapore. 2018 is when I started Tercep. My co founders went on to LinkedIn and Google to build some of their ad products. So one of the challenges that we faced a lot at Wizori was building what we call as this data Infrastructure, Right. So disproportionate amount of resources and money and people and time went into this. And when I moved on before starting through shift, one of the things I was thinking is that the same problem is obviously prevalent across the ecosystem. And that was my aha moment. Okay, why don't we build this so that helps the entire attic ecosystem move forward faster.
- [02:45]
A
Right. Okay, so what is it? What is data infrastructure? I assume you're mostly selling to companies with a lot of data. Right. So I need a lot of data. And now what am I getting when I, when I work with you?
- [02:56]
B
Which company in the world doesn't have a lot of data these days?
- [02:58]
A
Right, true.
- [03:00]
B
So when I say data infrastructure, what I mean is that all the tech and tools required to collect, process and visualize data, specifically here we're talking about monetization and marketing related data. Right?
- [03:13]
A
Right.
- [03:13]
B
When you look at media companies, there's a lot of monetization data, editorial data, analytics data that gets generated. If you look at an ad tech company, there's a lot of log level data. Many times they have an in house ad server, there's data sitting in third party platforms, et cetera. Right. So the ability to collect all of this efficiently, at scale, ability to process it and create visualizations, this becomes a huge challenge and requires a lot of resources and investments from these companies. That's what I mean by data infrastructure.
- [03:42]
A
Right. So but if we want to break it down further, right. Let's say I am ingesting log files. They show up, I guess in S3, and now the goal is to get them into Snowflake or Redshift or another database. I assume what you're doing is getting from point A to point B.
- [03:59]
B
Not necessarily. That's one part of the problem. To go from log files to getting insights out of log files is a lot of work to be done. What Thirdshare promises is to take you from an S3 bucket where you have all your logs to power BI or visualizations, fast visualizations within, I would say, days. That's the promise that we bring to the table. Everything that's in between, whether you want to take it to Redshift or you put it on a data warehouse or Snowflake or do something else, build a pipeline, optimize all of this stuff, etc. Is managed and offered by Terceft, so promises logs to dashboards.
- [04:40]
A
Is it in your cloud or if I want the data to show up in my data warehouse, do I use or is Tercep going to your data.
- [04:46]
B
Warehouse, so it can go to our data warehouse if you subscribe for the full stack solution if you want. So our offering can be broken down into various modules. And if somebody wants to subscribe just for getting data from APIs into their data warehouse, or do some transformation and then use it, those options are available as well.
- [05:04]
A
I see. How do you bring in third party data? If I want to join my data.
- [05:08]
B
With other sources, we have over 400 different API integrations right now with pretty much every adtech, martech company, all the SSPs, ASPs, AD servers, AD changes, AD networks, as well as the analytics companies, the CRM company, Salesforce, and so on and so forth. Right. So any of these data sets that your team is working with, we can connect with an API integration, click of a button, authenticate, get the data onto the ship, transform it, normalize it, and from there, you know, combine it with your logs, combine it with third party logs. Sometimes you're working with trade desk logs or DV360 logs or GAM logs and so on and so forth. So we get all of that data, combine it, normalize it, transform it, and then we give it in your data warehouse. Or we could help you visualize it on our platform as well.
- [05:51]
A
Right. So what's the technology lift of the customer or how sophisticated are the customers who generally want this kind of solution?
- [05:58]
B
So as I was saying, there are two kinds of companies. So one is companies who don't want to invest any technical resources or infrastructure to set this up running. And so they get a full stack white labeled fully managed, offer zero investments. From a technical perspective, there are companies who want to move faster with their BI plan, who want to accelerate their infrastructure or do want a better infrastructure. And they want to focus on optimization or machine learning or algorithms and stuff like that. Obviously they need to have a BI team. But what we promise is get all of this data, process it, make it available in your data warehouse, and then your team takes it forward from there.
- [06:36]
A
Right. And for those more sophisticated teams, is this like a complete service where you're doing all the deployment, or can they also kind of tweak the way aggregation works or the way the joins work?
- [06:48]
B
They can, they can. So one of the things over the years we've built a lot of internal tools and processes to allow for a lot of customization changes of joins, logics and stuff like that, which our customers leverage. And so they can set a bunch of rules of how the data needs to be processed before it needs to be.
- [07:06]
A
Right. Some of the tricky parts about dealing with this much data are things like errors and retries and catch ups. So how does your system handle all of that?
- [07:16]
B
Like I said, so we've built a lot of internal processes and tools to send alerts internally, to have exponential retries when things are not working, to pinpoint who is supposed to look at issues when it comes up, how our customers need to be informed, at what point in time who's going to get in touch with Xander if their API is breaking and so on and so forth. Right.
- [07:36]
A
Not to pick on Xander. We love Xander. Right, right. That makes sense. So why don't you take us through like a use case, especially for maybe you know, like a media company. Is it, you know what, where have you had successful deployments that would be interesting for our listeners.
- [07:53]
B
So a large news publisher based out of the US the top five news publisher based out of the US Pre tercept Everything was manual, everything was all Excel sheets. They had close over 150 different Excel sheets. They were trying to build out data studio dashboards. It was breaking all the time. They tried to set up the bigquery struggling and stuff like that. Right. So they went from there to a smooth setup where they have data coming in from about 50 different sources. They have data coming from the GAM logs, they have Google Analytics logs, they have data coming from 30 different SSPs that they're working with and so on and so forth. It took us about a month or a month and a half to get to an end to end, know, working sort of use case and get to a point where they were comfortable with it. But from there, right. They have now dashboards for their editorial teams, they have dashboards for their ad ops team for their leadership team that's automatically updated on a daily hourly sort of basis and gives them a lot of granularity in terms of data and visualizations and stuff like that. Right.
- [08:49]
A
What's the skill level of the people who are the customers in that area? Are they like business analysts, product managers, you know, or are they, are they developers themselves?
- [08:58]
B
Mostly ad operations folks, product managers and leadership. Right. Monetization leadership teams.
- [09:05]
A
Right. And can they give like dashboards to less sophisticated people like sales. Sales executives and people not to disparage sales executives, but still they could.
- [09:14]
B
They do. They do actually. Right. So one of the key factors that one of the key features that we've built out is what we call as role based access control. So you can drill down to every individual metric or dimension or what level of data exposure that you want on this particular login. So our customers do end up creating a lot of different logins for different team members. Many times they have divisions, they have a US division or a programmatic division. And they don't want to show direct data to programmatic or whatever. Right. So all of that is pretty easy. Right. And they don't want, for example, sales folks to have edit access, they don't want analysts to have global access. So all of that is super easy to set up.
- [09:53]
A
Right, right. Okay, take us through the commercial model. I assume it's not free. How are people paying for this?
- [09:58]
B
So we essentially charge for ingesting data and we spoke about two kinds of data sets that we ingest. One is log level data sets and two is aggregated data sets through API integrations. For log level data sets, we charge a flat monthly fee linked to the volume of records that we ingest. This could be auction records or bid records or impression records, and so on and so forth. On the API integration side, we charge a flat monthly fee for the number of data connectors that are feeding data into two shift. We start with a minimum of 10 data connectors and scale it from there.
- [10:30]
A
Okay. So the part of the charge is 10 data connectors. Which means like if I have a new data set, it's like a la carte pricing above the volume.
- [10:38]
B
That's right, yeah.
- [10:39]
A
And what about consulting or setup? It seems like there's gotta be a lot of work of getting the data flowing in the first instance.
- [10:47]
B
We typically sign multi year contracts, so we're happy to invest early on resources to set it up, keep it running and things like that. And we've ensured that our pricing covers our ongoing consulting, maintenance, support, servicing.
- [11:02]
A
Got it. In terms of the competitive set, you know, is it mostly replacing in house development or there specialized vendors that aggregate data that sometimes you might be going head to head with?
- [11:13]
B
80% of our new customers have in house setups and they want to sort of move to a better setup. Sometimes we compete with other companies as well, depending on the kind of use case. When we are working with log level data sets, we compete with a company called Rhyldata, if you've heard of them.
- [11:29]
A
Yeah, of course. We had Michael Driscoll on the pod a while back.
- [11:33]
B
Fantastic, fantastic. And when we work with APIs and aggregated data sets, we work with depending on the use case, again, we compete with companies like Stack and Adjuster, Bert Intelligence.
- [11:44]
A
Yeah, that makes sense. All right, well, let's move on to the lightning round. So I'll ask you relatively Quick questions. You give me relatively quick answers. So what is your number one competitive advantage?
- [11:57]
B
Our number one competitive advantage is a combination of the number of use cases that we support. Oftentimes we just not just focusing on unified programmatic reporting or their campaign reporting. It's a lot of different use cases for a publisher. It would be comprehensive revenue reporting and stuff like that, E commerce reporting and editorial reporting and so on and so forth. And the reason why we can support a much wider set of use cases compared to our competitors is because a lot of customization and the way we can support our customers as well.
- [12:26]
A
Got it. And what is your biggest challenge?
- [12:30]
B
The biggest challenge right now is getting word out in the market. Right. So we are relatively new in the US Market compared to some of the competitors that I spoke about.
- [12:40]
A
So.
- [12:41]
B
So we are working hard. We believe 2025 is a pivotal year for us for penetrating the US market.
- [12:46]
A
Well, it's great. We're publishing this the first week of 2025. So I asked all the interviewees this question. So why won't, like Amazon and Google just crush you eventually?
- [12:58]
B
I don't think Amazon or Google are ever going to be working with thousands and thousands of other ad tech companies. Right. So they have a strong presence. But there are thousands and thousands of other adtech companies. Fragmentation is only increasing. You would hear of a new ad format or a new ad channel or a new ad company, ad tech company coming up every other week. Right. And Amazon and Google are never going to consolidate all of this. So it's just not a use case that they would probably be interested in. They would probably be interested in a full stack, end to end, walled garden sort of use case. So this is playing outside of what Amazon or Google would ever be interested in.
- [13:31]
A
Gotcha. Okay, last question. If Tercept was an animal, what animal would it be?
- [13:36]
B
I have heard this question from your previous podcast.
- [13:40]
A
It better be a good answer.
- [13:42]
B
Absolutely. And we did give it a lot of thought. And the thing that comes to mind most often is to look at like an octopus. An octopus being able to gather all of this data from various flexible tentacles and becoming the source of truth or the brain behind all this data that's getting gathered. So I believe an octopus is a good representation of the set.
- [14:03]
A
That's a great answer. Well, Gaurav, thank you so much for being here and telling us about Tercept.
- [14:08]
B
Thank you. Thank you so much, Ari. It was a pleasure. Thanks for listening. New interviews are added every week at marketecture tv and your favorite podcasting app.
- [14:27]
A
Thank you for listening to the marketecture podcast. New episodes come out every Friday and an insightful vendor interview is published each Monday. You can subscribe to our library of hundreds of executive interviews at marketecture tv. You can also sign up for free for our weekly newsletter with my original strategic insights on the week's news at News Market tv. And if you're feeling social, we operate a vibrant Slack community that you can apply to join@adtechgod.com.