A Data-First Approach to Marketing

Redpoint Ventures
10 min readOct 2, 2020


By: Travis Bryant

Even as B2B technology and sales and marketing infrastructure has evolved, a critical gap in visibility has remained. Namely, most businesses don’t know who’s hitting their websites — or how to best serve them. That frustration led former Salesforce CMO Kraig Swensrud and SVP and GM Sean Whiteley to build Qualified, the first conversational marketing application, which is taking a data-first approach to the problem by leveraging tight integrations across the sales and marketing stack from day 1. Head of Founder Experience Travis Bryant talked to Kraig and Sean about how they think about data challenges, why automation should be in service of your reps, and how any tool needs to help with two things: generating pipeline and driving revenue. This conversation has been edited for clarity and brevity.

Travis: We met and worked together a long time ago, when we were all at Salesforce. And you’ve spent most of your careers in B2B marketing. I’d like to take it back there, because I’m curious about that experience and how that informed the insights you had to start Qualified. What did you feel like the problem was that you’re focused on solving?

Kraig: This is a problem that we’ve been thinking about for more than a decade. When I was CMO at Salesforce, one of the things that people on the team used to tease me about was that I would run around the halls of the Spear Tower saying, “I’m blind.” What I meant is that I had no idea what was actually happening with demand generation in real time, and I owned the marketing-driven pipeline number for sales.

That was at the time where everything was starting to go digital, and everything we were doing would point back to the website. Even our Superbowl commercial said, “Go to salesforce.com to learn more.”

Then on the website, we had these forms that would capture emails. Those became the leads that fed the sales organization.

But I had no idea who was on the website in real time. It could be a tire-kicker from a country that we didn’t even do business in, or it could be a VIP or CIO of Coca-Cola.

We treated them all the same, and we didn’t even know who was there or what they wanted. In marketing we were clueless, and the sales team had no idea whatsoever.

I didn’t have this real-time visibility into what was actually happening in the website was the centerpiece of it all. That pain and that realization have stayed with us through our ventures after Salesforce. When we came back around to starting this company, we’re like, “Isn’t it absolutely crazy that all this time has passed, and nothing is different today than it was eight years ago.”

We were inspired by other smart messaging technologies that were evolving. Intercom was one of the companies that we looked at and said, “Wow, there’s like an Intercom inside of my SaaS product, and it’s really smart, and it can turn on and turn off based on all these real-time variables.”

That was the idea behind Qualified: The website needs to be connected into all of your corporate apps, especially your CRM, Salesforce at the core. It should tell you who’s on your site. What are they doing? What’s your relationship with this company? Are they a buyer? Are they a prospect? Are they an existing customer? Do you want to cross sell-them? What if you could connect the right person inside of your company to that individual in real-time to facilitate a conversation.

Our hypothesis — which our customers are starting to see come to fruition — was that you could generate more leads and more pipeline and move deals forward faster by having that visibility and the tools to take action.

Travis: Obviously the data that you use to enrich that visitor, to help drive a better conversation, gives the rep more insight about their demographics and behavior. That’s a key insight, but the collection of all of that data, if it’s not enriched or normalized correctly, can lead to garbage-in-garbage-out kind of situation. So how do you think about that challenge, where you make all the data you’re grabbing relevant for the human to have the conversation?

Kraig: It’s a huge integration problem that we’re solving. It’s enriching information. It’s de-anonymizing a website visitor. It’s reading marketing automation cookies. It’s querying data out of Salesforce, and traversing the Salesforce data model. And doing all that in 300 milliseconds, and presenting that data to the right human, so they can determine whether they want to have a conversation. Or, whether they even get alerted at all.

It’s not to say that you’re never going to run into a situation where you don’t make any mistake about who the person is, but with real-time data, plus some selling tact, you’re going to engage in more conversations with the right buyers.

The alternative is to do nothing, which is absolutely crazy. Or the old-school alternative is put a live chat on the website where you’re just talking to random people all the time.

Now that we’re two years into this business, we have good traction with our customers and have the data to show that it works in a meaningful way for companies in terms of pipe generation, revenue, the selling metrics that matter for companies, especially in the COVID era. Having the best access to your corporate data in real-time, and telling the right people, and sparking up those conversations, it works.

Sean: Travis, I remember we had this conversation early on when we talked about why B2B websites were really stupid. And, if you’re a photographer, you can use Squarespace — it’s got everything you need. If you’re a smaller retailer, Shopify is awesome and has really great e-commerce infrastructure inside of the website. But where’s the equivalent website for a B2B company? It doesn’t really exist.

We’ve always liked to do things that are six degrees of separation from the CRM. We started to kick around the idea of what if the website was connected to your CRM database? How would that change the way that you function in the day-to-day?

So, if someone comes to your website, you could identify them. Are they a customer or not? If they are a customer, what’s their AOV? What’s their customer health score? Who needs to know? What actions should you take?

If they’re not a customer, what kind of a prospect are they? Do they figure ICP? Are they in a key industry? Who needs to know? What actions can you take?

The question for us then became, how can you use the data to derive some intelligence you can understand, like who should we be talking to, at what time? What should we be talking to them about?

The key to all of that of course, is connecting into all of these data sources, which is a big integration challenge. How do you let the right people inside of a company know what they should do when someone of interest is on the site? That’s where we landed.

Travis: It seems like one of the approaches that has started to become popular, but maybe the hype-cycle hasn’t delivered is, if you’re integrating all that data, then you throw a bot on top of it with ML to drive the conversation.

We’ve seen some ham-fisted approaches where these formulaic responses drive a poor customer experience.

I know the approach at Qualified has been, “How do we arm a human to have that conversation?” But as ML gets better, as you’re integrating more data sources, where do you see the opportunity to orchestrate what a bot might do with what a human might do, so that it’s seamless and not a poor customer experience?

Sean: First of all, we do believe that B2B companies selling a product, you’re going to pay six or seven figures for those sales conversations. It’s a nuanced conversation, and there are many of them. Think about the last time you bought something expensive. Did you want to talk to a bot or did you want to talk to a human? At the same time, we do believe that when you have enough data, historical data, there should be effectively a way for your bots to learn about what’s working, what’s not, and support your sales team.

We take the approach of thinking about all these different experiences all over your website. If you’re a target account and you come from LinkedIn, you’re going to get one experience.

If you’re a prospect that’s ever been to the website, you’re going to get another experience. We think that over time, there’ll be a learning model underneath all these experiences that can make some recommendations for you about, “Hey, people that come from LinkedIn’s graph and go to your pricing page, have engaged with this experience more frequently than they’ve engaged with the other one.”

You can see it starting to look a little bit almost like A/B testing, but at the end of the day, we still believe that the power of Qualified is really making sure that your humans are always engaged in super high-value conversations.

If you can just manage that basic level of orchestration, you can have a material impact just by doing that.

But over time, the automation piece, the bots piece, you want to train it in terms of A, what’s driving engagement? Then B, what’s actually working. We always like to give you the answer in the context of pipeline and revenue.

Travis: You mentioned it’s really an integration problem to help synthesize all this data. One of the core sources being CRM, of course. You guys are former Salesforce execs. Salesforce is an investor in Qualified. That’s a decision that you made from day one, to have deep connectivity into that platform. So how do you think about what part of the stack you want to own inside the larger CRM stack? And maybe can you share just a battle scar or two as a startup working with a giant company like Salesforce, to try to integrate tightly to get that value and how you’ve approached that.

Kraig: The very first thing we did is we said, “We want to be the best-in-class. We want to be purpose-built for Salesforce.” While some of our competitors in the space really targeted marketing automation platforms like Marketo or Eloqua, we had a different perspective. We want to take VIP visitors and connect them with humans. If that’s the first objective, and the most valuable objective for the people, then you’ve got to start with Salesforce. Salesforce is the master. Marketo’s not the master. Pardot’s not the master. Eloqua is not the master. Salesforce is the master for salespeople. Salesforce needs to be the heart of our entire system. When we took that approach, it just felt natural to us, honestly.

Salesforce needs to be the heart of our entire system. When we took that approach, it just felt natural to us, honestly.

Now, dealing with Salesforce as a small company, we’re lucky, because we’ve worked there and worked with Salesforce before. This is the third company that we’ve built in the Salesforce ecosystem. We sold our first company to Salesforce. We were executives at Salesforce. We were able to navigate the waters and we know the people that could help us knock down some of the walls when we ran into them.

It’s a huge advantage. We built our codebase a hundred percent on the Salesforce platform. Obviously our sales cloud, and our marketing cloud, and all of our Salesforce apps were funded by Salesforce ventures, which is at least helpful.

Being an enterprise player, we were able to navigate the security hurdles, and all of the compliance hurdles that are required to be a certified ISV, to get on the App Exchange, to be able to participate in those programs.

I think that while it might not necessarily be easy for an early-stage startup to get to the level that we’re at so quickly, we had that advantage, which was great.

Sean: Travis, you obviously have as much experience as we do at walking into a customer, understanding how their sales and marketing process works and mapping Salesforce technologies to it. You’ve done this a million times and that’s a big part of this job. As you’re walking into companies, they’re running their sales and marketing process like they did 10 years ago. It’s the same. So there’s been a lot of evolution in technology, in integration and all of these things, but people are still running the same process they ran 10 years ago. A big part of getting this product right is simply feedback cycles from customers.

Now we’re walking into these customers and you’re having these conversations with them about, “Tell me how you’re doing your handoff from marketing to sales. Tell me where you’re spending your demand-gen dollars. Tell me about the SLH your SPRs team will have with your sales team, and what kind of systems landscape are you using, and what are your sales stages?” You have these conversations, and what we’re doing a lot of is, it’s a lot of consulting. We’re saying, “You could probably speed this up and here, you could probably skip this step.”

It’s getting them to use the product, seeing what’s working, seeing what’s not. Collecting their feedback, getting it into the product roadmap, building that out, onboarding some more customers. I think what’s changing is, we’re seeing the larger customers now, bigger customers with more sophisticated global sales teams, territories, named accounts. What we’re finding is the bigger the company, the more value we can add, because when you get that $700K deal that you probably would have missed if you hadn’t been quick enough, that’s a material impact. That’s what we’re starting to find as we go up-market a bit.



Redpoint Ventures

Redpoint partners with visionary founders to create new markets or redefine existing ones at the seed, early and growth stages.