Podcast: Play in new window | Download
What if some of your most valuable marketing insights are already sitting in your CRM, email platform, or an overlooked spreadsheet? Tim Hayden joins us to explain how businesses can take control of their first-party data, connect the dots between disconnected systems, and use AI to make smarter decisions that drive sales, repeat business, and customer loyalty.
Your First-Party Data Is Hiding in Plain Sight
Somewhere in your Google Drive (or Dropbox, or Box) there’s a spreadsheet of people who entered a contest you ran a few years ago. You haven’t opened it since. I know this because I have one too. (Or twelve.)
That spreadsheet, along with your CRM, your email list, and every online order you’ve ever processed, is first-party data. And according to my longtime friend Tim Hayden, CEO of BrainTrust Partners, it’s some of the most valuable stuff your business owns.
Tim and I go back close to 20 years. He’s been a trusted voice on mobile, social, and AI, so when he says he’s focused on customer data now, I pay attention.
What first-party data looks like for a small business
First-party data is information people gave you directly. Tim’s list covers more ground than you might expect:
- Form submissions on your website
- E-commerce transactions and in-store purchases
- Email subscribers and SMS opt-ins
- Your CRM and marketing automation platform
- Event sign-ups and contest entries sitting in Dropbox, Box, or Google Drive
Each of those people bought, subscribed, registered, or asked for more information. “There is a certain level of consideration that you have with everybody who’s given you their data,” Tim told me.
He calls it the tea leaves. Wrapped up in that data are behavioral signals about who your best customers are, what they want, and when they’re ready to buy again. That’s the raw material for repeat business.
Data sovereignty and the problem with rented land
Tim defines data sovereignty like this: “Sovereignty to me is about not having to depend on third parties.”
For the past couple of decades, many of us have built on rented land: Facebook pages, Instagram accounts, TikTok. We control our content (sort of) but not where it goes, and definitely not the data flowing through those systems.
AI adds another wrinkle. Tools like ChatGPT and Claude know a lot about the internet and very little about your business. Tim’s concern is that when you upload customer or operational data to these tools, you can’t always say where it ends up, and you’re not getting it back.
To be fair, the major AI platforms offer business plans with stronger data protections, so read your terms before you panic. But the underlying point holds. If your customer data is what makes your business different, you want to be the one controlling it.
Your customers aren’t the national average
I’ve been using AI to build buyer personas for a while now, and it does a solid job. But those personas lean on third-party data and national averages. If you’re in a community that’s an outlier, or you’ve attracted a specific type of customer, a generic persona can steer you wrong.
Tim agreed and took it further. “No two car dealers operate in the same exact way,” he said. Every business has its own workflows, its own decision-making, and its own audience. The data in your systems “is totally unique to a business,” and sometimes unique to a single location.
So use AI for research, but feed it your own customer data when you can, so it works from your reality instead of everyone else’s.
Meet Rich, Richard, and Richie
Here’s where things get expensive. Tim pointed out that a company might have customer records spread across 15 to 50 different systems. In one I’m Rich, in another I’m Richard, and in a third I’m Richie. (Nobody calls me Richie, but your CRM might.)
That fragmentation costs real money. You advertise to the same person twice, print three direct mail pieces for one household, and never get a clear picture of who that customer is.
The fix is what the data world calls a golden customer record: one unified customer profile that pulls everything together. Getting there involves identity resolution, which matches up those scattered records so Rich, Richard, and Richie become one person again.
Tim described two routes. One is a data warehouse like Snowflake, Databricks, or the big cloud platforms. The other is a customer data platform (CDP) like Tealium or mParticle. His company also built its own AI-based option, the BrainTrust Intelligent Experience Platform (IXP).
He also offered a word of caution. “Most people just think CDPs are magical. We just gotta integrate everything and we’re off to the races,” Tim said. “No, not that easy.” You don’t dump all your data in at once. You start with your business priorities and bring data in a logical order.
Where open weight AI fits
Tim is a big fan of open weight models, and it’s worth understanding why. Open weight models, like Meta’s Llama family or DeepSeek, can be downloaded (usually for free from places like Hugging Face) and run in a private environment. You can train them on your own data without sending it to a third party.
The benefits are privacy, security, and cost. In Tim’s view, running a private open weight model costs far less than paying for enterprise licenses on the frontier models.
The limitation? This isn’t a weekend DIY project for most small businesses. You’ll need technical help, the right database setup (Tim mentioned vector databases like Pinecone), and likely a partner who’s done it before. If you’re a five-person shop, this may be a “next year” conversation. If you’re a multi-location business buried in customer data, it could be worth exploring now.
Your first step is a whiteboard, not a demo
“Honestly, the first step is an internal one,” Tim said. Get your team in a room and have an honest conversation about the systems you use, the data you collect, and the reports and decisions that come out of it. Then talk about where you want to go.
He recommends a whiteboard so everyone sees the whole picture and can erase and redraw as needed. Invite IT and your agency, too, since they’ll need access to prove what’s working.
Once you’ve mapped your current state, you’ll ask vendors better questions and avoid buying software you stop using six months later. (We’ve all looked at a credit card statement and wondered why we’re still paying for that one AI tool.) As Tim put it, “We get excited about shiny objects.”
Next steps
Here’s what I’d do this week:
- Inventory your data. List every place customer data lives: CRM, email platform, e-commerce, POS, SMS, forms, and those forgotten spreadsheets.
- Book a whiteboard session. Get marketing, sales, operations, and IT together for an hour. Map what you collect and what you do with it.
- Pick one use case. Repeat purchases, win-back campaigns, or better segmentation. Start there instead of trying to integrate everything.
- Check your AI terms. Know what happens to any customer data you upload to an AI tool.
- Look for duplicates. Pull a sample of records and count how many Richards and Richies you’ve got.
Your first-party data is already working for you in the background. It’s time to give it a promotion.
This post was drafted with AI assistance from my interview then edited and fact-checked by me. Thought you should know.
Transcript From Tim Hayden’s Episode
Rich Brooks: My next guest is CEO of BrainTrust Partners, helping brands grow revenue through first-party data and applied AI. He previously led marketing at Cygna Labs and headed the mobile program at Edelman Digital in North America. He’s advised leading brands including Ford Motor Company, Dell Technologies, and Walmart, and is author of The Mobile Commerce Revolution.
Today, he’s here to talk about data sovereignty, the first-party data you’re likely already sitting on, and how you can use that data to generate repeat business and referrals. I am happy to welcome back to the podcast Tim Hayden. Tim, how you doing?
Tim Hayden: I’m doing great, Rich. It’s always fun to see you. How far back do we go?
Rich Brooks: I mean, close to 20 years. I think I first met you when you were speaking at an event called, The Big Conference, here in Portland, Maine.
Tim Hayden: That’s right.
Rich Brooks: And then we’ve just stayed in touch over the years.
Tim Hayden: That’s right. I mean, it was The Big Conference up there in Portland, Maine. And if I remember correctly, the week after there was also a big event down at Gillette Stadium that Chris Brogan and Justin Levy had put together, the Inbound Marketing Summit. This is before HubSpot had their Inbound.
Rich Brooks: Right.
Tim Hayden: Yeah, I mean, those were the good old days. Spent a lot of time in New England back in those days.
Rich Brooks: Absolutely. When you said a big event going on at Gillette Stadium, I assumed you meant another Patriots victory, but I’m sure Inbound was also good
Tim Hayden: Well, I have little to no interest in how the Patriots are doing, so it’s all right.
Rich Brooks: Now, it’s funny because I was going to lead with we’ve known each other for a long time, close to 20 years, which is surprising given how good we both look. In that time, you have been a trusted voice on social media, on mobile marketing, on AI, and a lot more. What now has led you to focus on customer data and how companies are using it?
Tim Hayden: Well, I mean, if you go back to my days, I’m a recovering mobile strategist is what I like to tell people. And back then, from 2008 to 2012, ’13, the book, The Mobile Commerce Revolution, actually published in 2014, and there was one chapter in it that showed systems integration of what I thought it would be at the time. I really thought the CRM vendors would be the ones to be able to get that right. They didn’t.
But you know, it was one of those things that I really started to think about. How do companies deal with real-time data transfers that come from in the store back to headquarters? Because of the simple things that happen that are very human things. People buy things, they leave the store, it doesn’t fit. I bought a box of shoes and one of them is a size eight, one of them is a size nine, you know? And even with all the checks and balances that companies have, these kinds of things happen. This is why Amazon has made it so easy for you to return things, right? Because they know will happen.
But my point about data is that you want to be able to know things like this as fast as you can. You want to do it for the customer’s experience so that in a frictionless way, they can go ahead and transact with you in both directions. And you as a company, you want to know when things aren’t right. This just means that systems are connected, right?
I used to say the neck bone’s connected to the knee bone, right? When you really think about how the internet has connected us, social media especially put that on overdrive for us. Companies internally have got to do the same thing with their systems. They’ve got to get them tied together.
So I started really in 2015, 2016, while I was at Cygna Labs living in the Bay Area. I started to study a little bit about machine learning, and I started to study a little bit about this rising category of customer data platforms and what companies were doing. Not just stuffing data in the Cloud, but actually bringing it into environments where you could build what they call the golden customer record, which would mean all the data from you that may exist otherwise in a bunch of fragmented systems. Maybe 15 to 50 different systems in a company who’ve got your records, your Bob here, or you’re Rich here, you’re Richard there, you’re Richie there. and that creates issues.
It creates issues because you spend too much on advertising. You print too many direct mail pieces that are going to the same house when you’ve got records like that. So, that’s how it all started. that’s really how it started.
Rich Brooks: So before we get too deep into it, I mentioned in the intro the term data sovereignty. How do you define data sovereignty?
Tim Hayden: Well, sovereignty to me is about not having to depend on third parties. And when you think about the couple of decades behind us, most of business has grown through on rented land. We outsource our websites, we have Facebook pages, we’ve had Twitter accounts, we certainly have Instagram accounts, we’ve had TikTok accounts. I mean, we’re spread out in all these different places, and we don’t control any of those places, right?
We control our content to a certain degree, but we don’t control where it goes. We certainly do not control the data swimming through those systems. AI compounds what you don’t have control of because the most popular thing people are doing right now is using Claude from Anthropic or ChatGPT from OpenAI do things faster and cheaper. They’re using a system that is trained and knows everything that was on the internet, but hardly anything about your business except what’s on your website, right? So you don’t want to upload your information to these systems, because you don’t know where it goes. Nobody can answer that question, but I just can tell you that you’re not going to get it back whenever you do.
So sovereignty is about you being able to control your data. You’ve organized it, you’ve structured it, you’ve resolved it. So sovereignty means that you’re not having to share your data with anybody. It means you’re not operating on rented land. And it certainly means that you’re able to capitalize on open weight models, which are private, secure, and much cheaper to operate than if you were to continue to use Claude, for instance.
These are the things that we, as companies, if you want to protect your intellectual property, if you want to protect and make secure the customer information that’s been shared with you, your inventory information, your operational information, all of it, you have that opportunity to do that today, whether you’re a small business or you’re a very large enterprise.
Rich Brooks: You’ve talked a little bit about third-party data, and it’s really interesting because a couple years ago at the conference, I did a whole thing about living in a cookie-less future, which never came to pass. But the idea of first-party and third-party data, when you talk about first-party data, what does that include and where might we already have it without even realizing it?
Tim Hayden: Well, I would challenge you on the cookie-less thing because this is what Apple did with iOS 14 and giving us application tracking transparency, our ability to turn off all the apps on our phone, including Apple TV or anything you have on an iPad, to turn off their ability to crawl your device, to follow you from the website to your email, back to the app. You don’t have a cookie or a pixel on your website now that is following people everywhere that they go the way that you used to. And that’s mainly because the websites and some of the device manufacturers now have made it very easy for the consumer to control whether or not they’re tracked or not.
First party data is all that data you’ve captured in forms from your e-commerce transactions or your cash registers in a store. You’ve got first party data, which is in all the places that are your owned or your hybrid. Again, your Facebook page and some of the other things that you may be doing with social media, some of the things you may be doing with aggregators if you’re listing yourself in third party markets. You do have an opportunity to capture as much of that data as you possibly can. Then you have your email, and if you have a mobile app that’s out there that people are using, or you use maybe a mobile marketing system for SMS text messaging, you have all that data. So your CRM’s going to have a lot of that data.
Your marketing automation system’s going to have a lot of it. If you’re using Dropbox or some of these other systems, Google Drive, I bet you probably have, and I’m saying this collectively to the audience, but I bet most people have got a list of customers that attended an event, that entered a contest or a sweepstakes, some other thing where you just had a turnkey collection of data, and you probably have a spreadsheet that you’re keeping somewhere in one of those online Cloud folders is what we’ll call them.
So to me, this is gold. These are people that have invested in you. They’ve opened their wallet to you. They spent their time with you. They followed you. They subscribed, they asked for more information. There’s a certain level of consideration, maybe not 100%, but there is a certain level of consideration that you have with everybody who’s given you their data, whether they did so through a transaction or they did through something that was more passive, like an email subscription. And to me, these are the tea leaves, right?
Because wrapped up in all that data is behavioral information. It is directional information some of the systems that have made mention of, and we can talk a little bit more about, these are systems that were meant to decipher all these signals, to give you the heads-up insight. Not to give you metrics, but to give you directional guidance on what to do next, what to do with your content, or which audiences and which people belong to that audience, and what do you need to say to them and when do you need to say it and how you need to deliver it. I mean, that’s all wrapped up in that first-party data. That’s why I’m so excited about it.
Rich Brooks: I agree with you. And one of the things that I’ve noticed is you can use AI as an amazing research tool to create a buyer persona. But all of that is based on third-party data and national or global averages where you may live in a community that is an outlier, or you may just have attracted a specific type of customer. And if so, you’re building your buyer persona off of the general persona instead of leveraging some of that first-party data that you spoke of. Then you’re really missing the mark.
Tim Hayden: Bingo, Rich. No two car dealers operate in the same exact way. They have unique institutional, maybe even tribal or cultural ways that they go about workflows, their decision-making, and certainly the corpus of their audience, right?
What we mentioned earlier in terms of that data that exists in your current systems, that data is unique. It’s like DNA. It is totally unique to a business, or if you have multiple businesses in multiple locations, it may even be unique to that location. People behave different. They use different language. Semantically, they talk different sometimes. You need to be able to understand the value in all that, and you need to be able to leverage that and capitalize on it.
Rich Brooks: So you mentioned a few of these places where some of our first party data might be stored. It could be in spreadsheets, it could be in the Cloud, it could be on a CRM or MailChimp or Gravity Forms. What would you recommend as the first start, a first step to kind of pulling it all together and making it useful?
Tim Hayden: Well, I think the first step is just being able to find the right size system for you to do this. I mean, I would say you got two directions you can go. You can go the data warehouse direction, companies like Snowflake and Databricks and obviously AWS, Azure, GCP, have got very affordable ways for you to be able to put all your data in one place. And then within those systems, a manner of what we call resolution. Of being able to combine disparate customer information into one record. They have the ability to do that.
On the other hand, you have customer data platforms, and these are companies on the enterprise side of things like Tealium or Hightouch or AI. and there are another hundred or so, but those are three that we’ve got quite a bit of experience with. mParticle’s another one we’ve done some work with. And then we built the BrainTrust IXP, the Intelligent Experience Platform, as an AI native system, an AI-based system, private AI, open weight, not a wrapper of ChatGPT that does this for you.
And I think to answer your question, even a step before that, it’s just think about the use cases and think about your business objectives first. Be able to illustrate those, what I think are directional framework for how you’d go about doing this, because you don’t want to put all of your data in there all at once at one time. I mean, you can, you’re going to have a sequence of which you want to do this based on the priorities for your business. And then that should give you the direction on where you would start for which data you need to do this with first, and then what data needs to come after that, and so on.
I mean, that’s really what I think you need to do. Whereas most people just think CDPs are magical. We just got to integrate everything and we’re off to the races. No, not that easy.
Rich Brooks: All right. You mentioned open weight AI. Can you define for us what’s open weight and why is that better than closed weight?
Tim Hayden: Or than the frontier models, right? These are systems where you don’t have any control over them. This is part of the differentiation between the frontier models and open weight. With open weight, you’re actually able to train a model only on your data. When we talk about the weights of the model, talking about the manner in which the model can conform to your workflows. You want open weight models, and these are models Meta’s family of Llama models or DeepSeek or Inkling or Meta just came out with another one called Muse, not the personal agent that goes on your phone.
Bottom line is there are companies, these models are free, you can get them via Hugging Face or through GitHub. Most of them have websites that will drive you to where you can download them, along with all of the instruction on what to do once you have. And there are companies like Pinecone, which is a vectorized database you don’t just put a model by itself within a Cloud, and then it goes to work. You start to have an issue with how it’s able to and how fast it’s able to act.
So vectorized databases actually put you in a position where a model’s able to do the left turns, right turns, the bouncing around, if you will, in terms of its behavior and how it searches for answers, and to do it much faster. So those are just a couple of the different vendors that are out there that are doing this.
And the point I’m making here is that open weight, because it’s controlled cost, to use these models is not near as much as it would be if you were to try to wrap or use an enterprise license for, say, a Claude or ChatGPT.
Rich Brooks: How are you approaching these challenges with BrainTrust’s intelligent experience platform, and who is your core audience? Who are you working with to help through these problems?
Tim Hayden: Yeah, I mean, the core audience is inclusive of upper-level management and executives, but also the middle and the meat of organizations, directors and managers who are the front line of operations, whether that’s customer operations, customer experience operations, marketing, fulfillment, right? Inventory management and fulfillment, especially for e-commerce organizations.
We love to get all these people in the same room, and when we get them all in the same room, have this incredible opportunity to hear different use cases and different objectives that they each have, but they’re all working with the same data. And they all have an opportunity now to have AI that’s configured to be non-redundant, with the outcomes or the service that it provides to people in different business units or in different disciplines of the business. So we almost always are inclusive of a CTO or CIO, or if it’s external to your organization, your IT services, managed services provider.
Usually we invite them to be a part of the conversation. If you have an agency you’re working with, absolutely they need to be involved because they need at least privilege access to this data, because you want to prove what’s working, what’s not working when it comes to advertising, which usually comes down to sales, right? This is just a huge opportunity because of AI and machine learning before it, you’re able to reconcile and correlate the information or the insights, the directional signals that come from otherwise siloed and separate operations.
When you think about inventory, fulfillment and shipping, supply chain, and then you come back over and you think about audiences or unique audiences that you want to advertise to or you want to market to or communicate with, you’re able to now start to have AI, within the IXP at least, give you richer insights and deeper insights because they’re able to see how these things affect each other. Ultimately, how are they affecting the top and bottom-line revenue activity? Which I think is at the end of the day, what everyone is measured on or should be measured on, is our sales up and to the right? Is our profit margin increasing, if not our total volume of sales increasing?
So I think as we have more agent activity on the web and we have less people using very linear search, these are things that we need to have be team sports because we need the power of observation and feedback from humans, especially as they’re trying new things. I mean, this is a first rodeo for everybody in business today.
Rich Brooks: Well, I think you’ve made a strong case for the value of first-party data and consolidating it into a way that we can actually use it and learn from it.
If somebody wants to get started today and get a better handle of their first-party data based on what we’ve talked about today, what do you think that first step might look like?
Tim Hayden: Honestly, the first step is an internal one. It is talking amongst your fellow teammates, it is having an honest conversation about the systems they use. Because that’s an easy one. What are you using and what are you doing with it? What kind of reports? What kind of decisions? What kind of forecasts? What kind of analytics are you running?
And then beyond that is probably just a conversation of what kind of data are you collecting? I think if organizations can take that first step and just kind of map that out to say, “This is our current state, this is what we do,” you’re going to be in better shape, I would say, fantastic shape to talk to any vendor. Because before they do a demo for you, you’re going to be able to tell them, “Hey, this is what we figured out in terms of how we do things and what we want to do differently.” It saves everybody time, and it also prevents you from buying software that six months later you’re just not using the way that you thought you would.
I mean, everybody’s had that happen. Everybody does that, especially when they look at their credit card statements and realize paying for three or four subscription things that maybe they haven’t opened or visited or used in six months or longer, right?
So, we get excited about shiny objects. AI especially, it gets us excited and we have to go and we have to subscribe to things. But at the same time, we did so many times without having that diligence on the front end. What is it we do today and what is it we want to do different? And what are the outcomes that we need to change, or what are the outcomes we’re trying to achieve? Do that exercise internally. Facilitate it amongst yourselves or bring somebody third party in sometimes. That might be the easiest thing. With a whiteboard, the old-fashioned way. I still don’t think there’s any better way for people to learn when they’re in a place where things are all in front of everybody to see at once. And your ability to erase and make the line go to the left side instead of the right side if you need to, it’s about illustrating and re-illustrating what it is you do to be sure you all agree on that. And then secondly, where do you want to go next?
Rich Brooks: Awesome.
Tim Hayden: That’s the first step people have got to take.
Rich Brooks: Tim, if people want to learn more about you, about BrainTrust, about the Intelligent Experience Platform, where can we send them online?
Tim Hayden: Well, they can visit braintrustpartners.ai. it’s a long domain, I’ll give you that, but it’s braintrustpartners.ai, and find me on LinkedIn where I’m active commenting on a lot of other people’s posts and posting a lot of our own content. I’m doing a lot of it right now for our BrainTrust AI Blueprint Builder, and for the IXP, highlighting some of the stuff we do along our partners, folks like Snowflake, Telium, some of these others that I mentioned.
Rich Brooks: Awesome Tim, thank you so much. Always appreciate hearing your take on everything and thanks for swinging by.
Tim Hayden: Thanks, Rich. I appreciate you having me, and I hope to see you soon somewhere.
Rich Brooks: Sounds good.
Show Notes:
Tim Hayden is the CEO of BrainTrust Partners, where he helps businesses turn first-party customer data and applied AI into smarter decisions, better customer experiences, and revenue growth. Learn more at braintrustpartners.ai or connect with Tim on LinkedIn.
Rich Brooks is the President of flyte new media, digital agency offering branding, web design, and digital marketing located in Portland, Maine. He is also the founder of the Agents of Change. He’s passionate about helping small businesses grow online and has put his nearly 30 years of experience into the book, The Lead Machine: The Small Business Guide to Digital Marketing.