Hacker News new | past | comments | ask | show | jobs | submit login
Launch HN: Freshpaint (YC S19) – Automated, Retroactive Alternative to Segment
57 points by malisper on Jan 6, 2020 | hide | past | favorite | 44 comments
Hello everyone!

We’re Fitz & Malis, the founders of Freshpaint (https://www.freshpaint.io/). We make it dead simple to collect data from your site and implement tools in a single click and without code.

With our Autotrack javascript, Freshpaint automatically instruments your site. Then you can use our point-and-click interface to visually create events after the fact, and send that historical data to any of your analytics or marketing tools.

Customer data powers nearly the entire analytics, growth, and marketing stacks of today’s B2C and B2B businesses. But what’s crazy is that customer data is captured and managed using technology that is literally over 25 years old: Let’s say you want to see how many users clicked your signup button or played a song. Or you want to take the users who added an item to their cart and engage them in a marketing campaign. You have to manually write code for each and every event to capture and log that behavior. Then you have to send to it your marketing and analytics tools.

Segment is great because it allows you to collect data once and send it to a bunch of different tools, but Segment is still built on legacy technology. It still requires a massive engineering effort to stand up and maintain. Just last week someone told us that at unicorn-level grocery delivery company, it takes 2-3 extra weeks to ship anything because of this tax.

Freshpaint automates the challenges here so that engineering can focus on building product, implementations can be done in minutes not weeks or days, and you can make your stack work for you instead of the other way around.

We were inspired by a phenomenon we saw while working with customers as early Heap employees. We saw that lots of customers loved Heap for product analytics, but that was only a small slice of the stack. Across analytics, product, and marketing it’s common to see a dozen other tools that require customer data. We kept seeing the same data pipelines and integrations built using the legacy approach of manual code.

We built Freshpaint to instrument your entire site and hook up your stack in just a few clicks, without having to write any code.

How it works:

1. Install Freshpaint’s Autotrack snippet on your site. It takes 60 seconds, and from that point Freshpaint collects every click, pageview, etc.

2. Connect over 70 destinations like Google Analytics, Amplitude, Hubspot, Fullstory, Intercom, and a data warehouse. This is done by copying and pasting an API key or account ID. See our integrations here: https://www.freshpaint.io/integrations

3. Create events in Freshpaint. Our UI provides a point-and-click way of creating events. Data is retroactively available back to the day you installed Freshpaint, regardless of when the event is created.

4. Send data to any destination in one click. You can even backfill historical data.

5. Use all that time you saved to ship something new, build an awesome feature you can upsell, or plot your next marketing campaign to take over the world.

Great technology is not just about product. We commonly hear from Segment users that it’s really expensive for what you actually get. Search HN and you’ll see tons of complaints and even horror stories about their pricing like this comment here: https://news.ycombinator.com/item?id=19221872. With all the work that’s required after you purchase, we agree! On top of automating all that work, our pricing is designed to be much more friendly.

We look forward to feedback from the HN community!

I remember evaluating Heap (granted, a long time ago, so I'm sure it's evolved a lot since then) and the auto-tracking capabilities and retroactive analysis sounded awesome, but in practice it wasn't as valuable or easy as I thought. I found that in general, knowing the questions you want to answer and structuring your tracking around that to be far more useful, and that actually being able to take mountains of generic behavioral data and making sense of it wasn't that easy and a ton of work in and of itself. E.g. retroactively finding clicks to a single 'signup button' is easy, but a lot harder when that button is on many different pages and you want to group them or segment them in certain ways, or the implementation changes in some places, etc. In the end, it seemed like it was doing engineering work just to parse the data back out so it seemed more sane to structure in a better way ahead of time.

I'm curious what steps have been taken in the past several years to improve on this and how Freshpaint is approaching this. I think it's a really interesting problem and if done well, could be massively valuable. Good luck!

We have been thinking about this a lot in my company (RudderLabs - where we have built an open-source product around auto-tracking). Totally agreed that managing/using all this auto-track data can be really overwhelming and sometimes it is just easier to just generate the events via code. Your maintainability point is also very important - with events generated via code, you have a single owner for updating the website and updating the events as opposed to auto-track where the responsibilities are split between website developer and marketer.

However, there are interesting use cases that are only possible via auto-tracking. Let's say a particular link on your website is broken (e.g. takes to an old sign-up page which is broken). You will be able to identify that that broken click "automatically" because the ratio of successfully signed up vs not-signed up users would be very low for people who click on that vs those who don't.

You can automatically highlight many such interesting insights by doing some basic stats counting. If you throw more advanced MLish stuff, you can find really interesting (but sometimes hard to explain) insights.

Happy to chat more if you want. Email in profile.

There's two main reasons you would want to use auto-track. You should use auto-track if:

- You don't have the resources to setup tracking code.

- You want to look at data that you forgot to track previously. We've heard a lot of stories of people launching a new feature and forgetting to setup tracking on it.

If you have the resources to figure out and instrument everything you want to track ahead of time, then you should go ahead and do so. A manually tracked implementation is more likely to be correct than a autotracked one. Even so, auto-track is still useful as a tool for when you have new questions you want to ask or you realize that you need data that you aren't currently tracking.

You should not use autotrack to perform fishing expeditions where you just look at all the data and try to make sense of it. You should still have some guiding questions and autotrack is a tool that makes it easy to get the data you need to answer those questions.

Such a cool idea. I got a deep dive Postgres training from one of the their founders at my company and he was insanely knowledgeable. We currently are having these problems with Segment as well. I'm going to try and push to see if we can switch off of this from Segment. Congrats on starting a company that is solving a crazy pain point. I saw something saying that you need a minimal-loveable product nowadays and this is it!

Perfect timing as we have a huge need for this for an upcoming launch.

Question...will you be able to replay/backfill as you add new integrations? For example, Klaviyo is a huge hole in your integrations right now especially for marketing and ecomm. If I install Autotrack now, set up a few events and then a month from now you add Klaviyo, I can just replay everything into Klaviyo?

Also FWIW, I was an early user at Segment (was in their seed deck as case study), Klaviyo, and many others. I'm a technical marketer so can be helpful as beta tester and user research. reach out directly if you'd like to connect michael@futuremood.com

Congrats to Freshpaint on your launch. The product looks super-cool.

Disclaimer: I am the founder of Rudder (https://github.com/rudderlabs/rudder-server/), an open-source Segment alternative.

We have the feature that you are asking for. We have an auto-tracking JS SDK which captures all events and dumps into S3. You can map these dom-level raw events to virtual events via a simple javascript transformation function. We also have replay so you can retroactively do the mapping and send old events to different destinations.

What we don't have is the nice UI to create virtual events from the raw events, that Freshpaint has built. Looks super cool!!

> Question...will you be able to replay/backfill as you add new integrations?

Yes, absolutely! As long as the destination is able to accept timestamped data, you'll be able to replay all historical data since you installed Freshpaint.

Totally agree with parent on the Klaviyo integration being a miss for ecomm. That and Mailchimp would help quite a bit.

I also use Segment with multiple clients that I work with. Will check this out as there's room for alternatives.

Ah ha! Your quoted "horror story" comment is mine, so I guess I'm obligated to check Freshpaint out. Will do :)

You mentioned that you were early Heap employees. At one point it looked like Heap was heading in this direction as well (with their own warehouse & destinations and sources, etc.) But I just checked their site out and it looks like they're back to going all-in on "capture everything ananlytics". Is that accurate?

So I suppose it would be fair to say Freshpaint:Segment :: Heap:Mixpanel (et al)?

Heap has changed directions a lot over the last several years. When I about a year ago, they were going all-in on product analytics. They were focusing a lot on new kinds of analysis features. The last project I worked on was their path analysis feature[0].

> So I suppose it would be fair to say Freshpaint:Segment :: Heap:Mixpanel (et al)?

Yep! You hit the nail on the head.

[0] https://heap.io/paths

We use freshpaint to connect to Amplitude. The setup was super fast and we had product analytics set up within minutes! The auto-track is a life-saver so that we can confidently ship a feature and then come back and set up product analytics after the launch, while knowing that we aren't missing any old data. The team is very helpful with a ton of experience in the analytics space, so rest peacefully knowing you're in good hands :)

This is a really cool product. Very impressive that you were able to build to many integrations with such a small team. Any advice on how you did that?


The secret is that most APIs that accept event data are pretty much identical. They have some method or endpoint where you send a JSON dictionary of properties. Once we had had the core event processing down, it wasn't hard to quickly add a lot of integrations.

Ah, interesting...

I've looked a bit at CRM integrations and they all do things slightly differently in ways that make it hard to factor out common pieces. Glad to hear that isn't the case in the event data space.

How many of these are real? I don't care how efficient you are as a team. I see 50+ integrations and one engineer on the team, on top of a complicated base product.

Color me suspicious...

Not vaporware! All of our integrations listed are all real and can be used today.

Similar question to the grandparent - wondering how you go about maintaining a relatively large catalog of integrations with limited resources (I worked at an integration platform as a service company and remember maintenance of those to be the most resource intensive part of the product)


How do you prevent collecting PII/passwords/other sensitive data to avoid situations like Mixpanel's Autotrack fiasco? https://mixpanel.com/blog/2018/02/05/update-autotrack-data-c...

We use Freshpaint for our site (heraldhq.com) and loved how we could backtrack analytics events. Excited to see where this goes.

Does the $300 a month pricing kick in at the first instance of 3,001 unique site visits in a given 30 day period?

No. We know that sometimes companies have random spikes that don’t correlate to sustained user growth. Higher pricing would kick in after the 2nd month of being above the volume limit.

> Tracking code is dead. Technology has evolved.

That is definitely not the case and this sounds like a cheap marketing wording on your web page. You may be visiting a prospect and may learn it from them the hard way how tracking code crushes "autotrack-and-collect-anything" approach by leaps and bounds.

> unicorn-level grocery delivery company

At this point you can just say Instacart :)

Very cool! Does this work on mobile as well?

Any time I hear about a hot new analytics company, my first thought is always: how do I opt out?

Freshpaint, how do I opt out of your data collection? Now, into the future, and also for any historical data that you might have collected about me.

We respect Do Not Track. If you have Do Not Track enabled, we won't collect any data from you. Unfortunately Do Not Track has a number of fundamental issues and I'm not aware of any better way to signal you want to opt out.

You really shouldn’t support DNT, it’s a useless “standard” that’s half-deprecated (Safari is removing support in 2020) and will potentially introduce significant bias into your customer’s data.

100% this. Hopefully by the end of the decade data collection will default to opt-in rather than opt-out.

No, it will just be all server side and you won't have any choice.

Good, server-side data is harder to fingerprint and track.

It’s pretty trivially easy to set a first-party cookie or ID for a user in local storage for non-logged-in users.

Of course for SaaS all product usage is logged-in so fingerprinting doesn’t really apply.

You can't set or read local storage from the server. As for first-party cookies, not only can those be blocked, as they don't let the server track across sites, which is the most annoying part.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact