Customer Story - Drivvn

[joethreepwood]

@kevan-gilbert and I are going to chat with John, from Drivvn, for a customer story. The main goal here is to get information on how they use group analytics, as this is a premium feature which we currently don't offer much info for on the site - leading to a high churn rate for it. Once we have a case study, we can then feature that into the onboarding email flow to encourage usage, and feature it throughout the site.

I'll lead the chat, with Kevan observing, so he's prepped in the future to take on additional case studies as they arise. Kevan: for your reference these chats are usually very conversational and the main goal for me is always to get the interviewee into a relaxed state so they feel happy to share practical examples. My go-to stance is usually to ask simple questions, parrot-back to people occasionally, and basically let them keep talking as much as possible. I record with Google Meet, send recordings to rev.com for transcription afterwards, then write those up into articles for posthog.com/customers

Standard sort of questions I'll ask below!

---

Tell me about yourself. What's your background?
And what about Drivvn - what do you do, what's your role?

How long have you been using PostHog?
How did you discover PostHog?

Were you using anything else before? Have you personally used other tools?
What PostHog tools do you use at the moment?
Did you always use this suite, or have you added more tools over time?

Your company is B2B. Is that why you use group analytics?
What type of info are you usually capturing at the group level?
How do you use that data? What sort of analysis are you doing?

I'm told you use groups with feature flags especially?
What sort of thing do you use feature flags for?
Why do these activities run as flags, rather than experiments or shipped features?

How has your PostHog usage changed over time?
What have you learned about your users and groups?
Are there any direct changes you've made as a result of this?

How are you sharing your discoveries and acting on them?
What's the biggest advantage you've discovered from using PostHog?
Is there anything we missed?

[kevan-g]

Thanks Joe!

Links for us:

• Customer website
• In Vitally
• In PostHog...many different instances and users with Drivvn-relatedness.

[kevan-g]

Hi @joethreepwood!

I've got a draft written for us to review: I'll paste it here, and it's also viewable as a Google Doc.

How Drivvn drives growth with group analytics and feature flags

What if shopping for a new vehicle was a user-centered, seamless, thoughtful, online experience — no dealership required?

I know, hard to imagine: the "car salesman" trope is perhaps the most notorious of them all, second only perhaps to the dreaded B2B SaaS sales approach. But what if it didn’t have to be that way?

That’s the future Drivvn has been building. Supporting a network of more than 60+ automotive ecommerce providers, Drivvn is the digital engine behind global vehicle brands from Ford to Fiat. Known for creating brand experiences that are design-driven, human-first and still do great at the “selling” part — you’d almost think it was magic.

Under the hood, it’s classic PostHog.

Well, that and the legend known as John Laramy.

Drivvn’s Head of Products Engineering, John and his team have been using PostHog to up their game for over two years. While his team designs and builds these online shopping experiences for car brands, PostHog is what helps them optimize and customize the experiences for end users.

To help you visualize it, as a user, let’s imagine you’re shopping for a Ford vehicle over on this site. Here, you’ll have an easy time comparing vehicles, and shopping for what you need — but what you may not know is that your experience is quickly being customized to better serve you, as you go along, all orchestrated by John & Co.

For example, through Experiments in PostHog, they can slowly control the release and test new changes: first, they test the feature with their in-house team at Drivvn. If it functions correctly, they’ll use PostHog to release it externally, to a small audience: say, 10% of visitors. If that’s successful, they’ll roll it out to 50% of users — a safely controlled, step-by-step version of A/B testing that ensures no single, wild change wipes out any of their hard-earned progress.

“It also helps us target a specific brand experience to the right audience,” explains John. “If we’ve identified that a visitor fits more in the business-to-business category, we give them an experience that matches their needs. The brand experience between B2C and B2B can certainly be different.”

But it’s wider than even that: as an agency-style operation, John and his team hold an ecosystem view. It’s not just one website, one audience base. “It’s also across different brands as well — you might release a feature for one brand, but not enable that across the other ones, because you don't know how well it's going to perform.”

Through all these tweaks and changes, John’s team can report to their own customers on how well their branded digital experiences are performing. Through dashboards, the reporting area helps them have those conversations — and it’s the data that is changing people’s minds.

“With one customer, I would say they were very against this [experimental, data-driven, A/B approach,” says John. “We then did one small-ish change — as subtle as changing the CTA on one page. When we showed the benefit of that, suddenly it was like a lightbulb moment. It clicked: ‘I understand the benefit of that now.’ After that, they were very keen to continue doing that approach.”

“One of the reasons we went with PostHog was because of the wide range of features,” says John. “We're using analytics heavily today across all those brand experiences. We're doing lots of feature flags, and also experiments…all towards helping people get to a clear outcome.”

As someone who has been in engineering roles for more than a decade, John cares about the details. “We’ve used other tools like MixPanel and Google Analytics, but I think the real thing that we were after was that it was third-party and cookieless.”

“We're big advocates of PostHog at Drivvn — it's not just myself,” John continues. “There's a number of other people who are in PostHog on a daily basis looking at that good stuff. The documentation is very good. You guys are obviously releasing new features often. When we do have questions, we use Slack and we're able to reach out to you, and we get a response in a timely fashion.”

What’s next for John and Drivvn? Pushing PostHog to the limit. “I think we are utilizing maybe 70% of the new features,” he says. “We've only very recently started to play with the data pipelines —it’s something that we'd like to play with in the future, as having these rich insights is really helpful.”

[joethreepwood]

Left some comments on the doc -- it looks great and gave me a few ideas for how we can take it to the next level. Want to take a look at the suggestions and turn it into a PR?

This old case study PR will point you in the right direction if needed. Let's just keep it as draft until we've got a final version to send back to John for his approval.

You'll also need to put an art request (like this) into Lottie for some blog art.

The two main things that are missing though are:

• A nice, big, box-quote. We have these on all the other case studies and it helps make it pop out. Optional, but recommended!
• We don't mention group analytics at all. We need to bring this in much more overtly, as it's the focus of the article. The intended use here is to put this case study in an email flow to support group analytics adoption, so we need to make sure that message pops out.

[joethreepwood]

Left some comments on the doc -- it looks great and gave me a few ideas for how we can take it to the next level. Want to take a look at the suggestions and turn it into a PR? This old case study PR will point you in the right direction if needed. Let's just keep it as draft until we've got a final version to send back to John for his approval. You'll also need to put an art request (like this <#9695>) into Lottie for some blog art. The two main things that are missing though are: - A nice, big, box-quote. We have these on all the other case studies and it helps make it pop out. Optional, but recommended! - We don't mention group analytics at all. We need to bring this in much more overtly, as it's the focus of the article. The intended use here is to put this case study in an email flow to support group analytics adoption, so we need to make sure that message pops out.

…

On Tue, Jan 21, 2025 at 11:56 PM Kevan Gilbert ***@***.***> wrote: Hi @joethreepwood <https://github.com/joethreepwood>! I've got a draft written for us to review: I'll paste it here, and it's also viewable as a Google Doc. <https://docs.google.com/document/d/1TgRlcDlq333BYzbE1DaZzDlxzP7hLbW55kc8PD-35ik/edit?tab=t.0> How Drivvn drives growth with group analytics and feature flags What if shopping for a new vehicle was a user-centered, seamless, thoughtful, online experience — no dealership required? I know, hard to imagine: the "car salesman" trope is perhaps the most notorious of them all, second only perhaps to the dreaded B2B SaaS sales approach. But what if it didn’t have to be that way? That’s the future Drivvn has been building. Supporting a network of more than 60+ automotive ecommerce providers, Drivvn is the digital engine behind global vehicle brands from Ford to Fiat. Known for creating brand experiences that are design-driven, human-first and still do great at the “selling” part — you’d almost think it was magic. Under the hood, it’s classic PostHog. Well, that and the legend known as John Laramy. Drivvn’s Head of Products Engineering, John and his team have been using PostHog to up their game for over two years. While his team designs and builds these online shopping experiences for car brands, PostHog is what helps them optimize and customize the experiences for end users. To help you visualize it, as a user, let’s imagine you’re shopping for a Ford vehicle over on this site <http://ford.buy.co.uk/>. Here, you’ll have an easy time comparing vehicles, and shopping for what you need — but what you may not know is that your experience is quickly being customized to better serve you, as you go along, all orchestrated by John & Co. For example, through Experiments in PostHog, they can slowly control the release and test new changes: first, they test the feature with their in-house team at Drivvn. If it functions correctly, they’ll use PostHog to release it externally, to a small audience: say, 10% of visitors. If that’s successful, they’ll roll it out to 50% of users — a safely controlled, step-by-step version of A/B testing that ensures no single, wild change wipes out any of their hard-earned progress. “It also helps us target a specific brand experience to the right audience,” explains John. “If we’ve identified that a visitor fits more in the business-to-business category, we give them an experience that matches their needs. The brand experience between B2C and B2B can certainly be different.” But it’s wider than even that: as an agency-style operation, John and his team hold an ecosystem view. It’s not just one website, one audience base. “It’s also across different brands as well — you might release a feature for one brand, but not enable that across the other ones, because you don't know how well it's going to perform.” Through all these tweaks and changes, John’s team can report to their own customers on how well their branded digital experiences are performing. Through dashboards, the reporting area helps them have those conversations — and it’s the data that is changing people’s minds. “With one customer, I would say they were very against this [experimental, data-driven, A/B approach,” says John. “We then did one small-ish change — as subtle as changing the CTA on one page. When we showed the benefit of that, suddenly it was like a lightbulb moment. It clicked: ‘I understand the benefit of that now.’ After that, they were very keen to continue doing that approach.” “One of the reasons we went with PostHog was because of the wide range of features,” says John. “We're using analytics heavily today across all those brand experiences. We're doing lots of feature flags, and also experiments…all towards helping people get to a clear outcome.” As someone who has been in engineering roles for more than a decade, John cares about the details. “We’ve used other tools like MixPanel and Google Analytics, but I think the real thing that we were after was that it was third-party and cookieless.” “We're big advocates of PostHog at Drivvn — it's not just myself,” John continues. “There's a number of other people who are in PostHog on a daily basis looking at that good stuff. The documentation is very good. You guys are obviously releasing new features often. When we do have questions, we use Slack and we're able to reach out to you, and we get a response in a timely fashion.” What’s next for John and Drivvn? Pushing PostHog to the limit. “I think we are utilizing maybe 70% of the new features,” he says. “We've only very recently started to play with the data pipelines —it’s something that we'd like to play with in the future, as having these rich insights is really helpful.” — Reply to this email directly, view it on GitHub <#10341 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AUA6UKPSBANFZSS4AEFL5GD2L3ND5AVCNFSM6AAAAABVE3MGVCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMBVHE4DCNJWHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[kevan-g]

Awesome. Thanks @joethreepwood

I've edited the draft to incorporate your suggestions, including adding some additional content about Group Analytics. Feel free to review the changes! I'd love to hear your thoughts on whether this does what we want it to re: the group feature.

I can see a few great options for pull-quotes, will add.

Art request here.

I began to do a PR for this (I have a drivvn.md page started in /customers locally; pull quote prepped), but realized there's a button on the website now that let's you do a new Post through a front-end CMS: https://posthog.com/posts/new -- but, pressing Post didn't visibly accomplish anything. Know anything about that?

Do you really send PRs themselves to external contacts for review? (My expectations would be that google docs would be easier for external link sharing and commenting, but perhaps this is how we lean-towards-shipping: "we're about to go live, y/n?" vs - "please leave endless comments.")

[joethreepwood]

PRs are preferable for collaboration. While Google Docs are fine for initial drafting, the work ultimately needs to be turned into a PR to ship. This also fits with the hierarchy of work: PR > Issue > Slack > Email. Sending Google Docs back and forth falls under email in this context and can be inefficient.

As discussed with John on the call, he’s comfortable reviewing directly on a PR. Comments can be left there just as easily as in a Google Doc, and most of the people we collaborate with are engineers who work best in this flow. Whenever possible, let’s always jump straight to PRs for feedback and collaboration.

I'm not familiar with that /posts tool so that's probably a question for @corywatilo -- but my assumption based on the fact that it's called Posts is that it'll publish to the Posts feed, rather than the /customers section. So, again, we should PR here.

[posthog-contributions-bot]

This issue has 2380 words at 6 comments. Issues this long are hard to read or contribute to, and tend to take very long to reach a conclusion. Instead, why not:

1. Write some code and submit a pull request! Code wins arguments
2. Have a sync meeting to reach a conclusion
3. Create a Request for Comments and submit a PR with it to the meta repo or product internal repo

Is this issue intended to be sprawling? Consider adding label epic or sprint to indicate this.

[kevan-g]

Great! PR is here: #10474
I've shared with John.

[joethreepwood]

Prefer to leave this open until the case study merges, please!