Hacker News new | past | comments | ask | show | jobs | submit login
Segment Is Now Free for Early-Stage Startups (segment.com)
248 points by brackin on Feb 21, 2019 | hide | past | favorite | 83 comments

I have to say, a few years ago I was the world’s biggest Segment evangelist. My absolutely favorite analytics service, and for any web project I worked on, installing Segment was step 1. I ran it primarily on my own company’s site (100k visitors/month, most of whom were not monetized) and I think I paid them $79/month. Worth every penny as a router to send up to GA, Heap, and others. Their data warehouse became a wonderful adjunct part of my analytics stack (and maybe my bill went up to $200 or so when they added that.)

Then they changed their pricing so that bill became $1200 (per month!). Total insanity. When I did some consulting work for another startup that had even higher traffic numbers (150k-200k uniques) but was also very lightly monetized, I actually lost some serious political capital with the CEO when he saw the Segment price quote of nearly $2000 per month. He thought I was trying to fleece them.

Sorry guys. I think you’ve built a REALLY awesome product and I / similarly small businesses are clearly not the target demographic anymore, and I’m sure you’re now making metric boatloads of money from big enterprise customers who don’t care about a $20k annual bill for analytics, but man, it’s disappointing to be so priced out of what was once a great part of my toolkit.

100% agree. We are working on a YC W18 company (www.cointracker.io) and loved that Segment made it easy to get up and running on a bunch of metrics platforms. Really slick product!

Unfortunately, the recent price hikes are completely out of line with the value we are getting with the product and we are being financially forced to stop using the service. Chatting with the sales team has been extremely painful and unhelpful. It has been incredibly validating to hear that everyone else here is feeling the same way.

@Ilya — I really hope you'll consider taking this feedback to heart. You have absolutely nailed the product. The pricing needs a complete overhaul.

Thanks Chandan, we're hearing this loud and clear. Open to give us feedback on pricing/packaging changes we're working on? If you're willing, can you send me an email at: ilya @segment.com

Just curious, what are the reasons to be on "a bunch of metrics platforms"? Specifically, what are the various problems you were trying solve with different tools?

One of the big benefits of Segment's higher tiers is that you can replay historical data into new tools, which I don't believe is typically possible if you just integrate a new platform directly. Business needs change.

In our case we route analytics to Google Analytics, Intercom, and a Postgres database (using Segment's warehousing support) - that gives us high level analytics for marketing, customer support tooling, and raw data for the team who work with that.

We provide a white labelled platform for a few enterprise customers as well, and Segment allows us to route data to the appropriate accounts for each of those services, rather than having to set everything up manually each time.

Hey, just wanted to let you know that we're reading/listening and this is really helpful feedback. I'd love to do a phone call, and get your feedback on a few pricing/packaging changes we're working on. If you're willing, my email is ilya @segment.com

Since you’re here, I highly agree. Segment is awesome and I’d love to pay for it but I need to track entire lifecycle for customers, including their original acquisition. That means tracking anonymous web visitors, and that immediately puts Segment into the “whoa, it’s how much?!?” category for me.

Very cool of you to be active on this Ilya! Happy to chat - I'll drop you a line.

I'm in the same boat. I love love love segment at Reforge (and I hear Segment loves Reforge, too). I've built some pretty awesome data pipelines (Personas are my favorite feature of all time) and have armed the marketing team to the teeth.

But my side project, Casting Call Club, has 150k uniques a month, most of which are not monetized. There's just simply no way I can justify paying for Segment. Many of the ideas I think about are simply lost because it would take too much time to build out myself.

We get almost 500,000 uniques a month, mostly unmonetized (though that's changing, slowly). I would LOVE to integrate an enterprise tracker but the pricing is just absurd. I mean, there's no way we can even 'try' it. We grew alot faster in organic traffic than we expected.

It's definitely meant for large enterprises, not for growing ones.

I guess, I'll throw my hat into the ring as well. Former customer here as well. Just finished pulling out all the Segment hooks from the code base for same reasons as others.

Similar experience here. When I first discovered Segment there was a free plan, the value was clear (install just one script + ability to replay), and there were affordable plans for small businesses. So I recommended using it right left and center.

I get the reasoning behind the price change, and I presume it was good for their bottom line, but it felt like extortion.

The current prices are still a far cry from what they were back in 2015 [0] and the most valuable feature (replaying historical data) now seems to be enterprise-only. It still looks like great value once you reach a certain size but I'd be hard pressed to use this new startup offering or recommend a client to do so.

On a more strategic note, the door seems wide open for another Segment-like startup to capture early stage businesses. The only thing that might be saving Segment in this respect is that Google isn't suggesting a competitor when you type "Segment vs" -- yet.

[0]: http://web.archive.org/web/20141231114510/https://segment.co...

Segment is best used with identified real users (i.e., you have their email), not anonymous web traffic, mainly because the former is where all the value is and the latter will unnecessarily drive up your bill.

That said, they could consider a cheaper, lightweight option that pipes traffic for you with a single tracker (the original value prop) but without the identify call and custom track calls, which is where all the downstream and up-market value is with respect to personas and the swing to become a CDP.

What was the thing you most liked about this service compared to its competitors. It's the first time I'm hearing about this service and couldn't really understand why people are raving so much about it. To me this looks like YAAS so I guess best ask someone who used it for sometime and paid $79/mo for it.

We had the same experience. Segment's pricing changes in 2017 were borderline fraudulent. We were automatically switched from the free plan to a plan that cost over $450/month without ever agreeing to it. Segment's "free" plan was a bait and switch, I wouldn't trust it.

Segment is a good concept but when they bumped their pricing up to ludicrous mode, it justified a small rewrite to reproduce their functionality in our back-end to drop them and we haven't looked back. No way i'll be going back to their service factoring that experience.

Interested to see what the back-end of this 'free period' will cost for the startups who get in bed with them - $120 per month for up to 10k MAU and then 'custom' pricing beyond that according to their pricing page.

It's worth a lot less than that to us.

Hey, wanted to say that I really appreciate the feedback and we're releasing this program exactly to address it. This is Ilya, one of the co-founders of Segment.

Re: how the program works on the backend — For new startups (<2 years, <$5M raised), the Startup Program provides $25k of annual credit (and $50k credit if you're applying thru an accelerator) on the team plan which makes "Segment Free" for 99% of the thousands of startups that use Segment. You can apply for it for multiple years, and after you graduate, we'll provide graduated discounts off our standard plans (50% in year 1). This is designed to be affordable for the entire time that a startup is getting off the ground.

Re: value over time — we see companies of 2+ years of age use an average of 7 source types (web, mobile, crm, stripe, ..) and 8 destination types (google analytics, event analytics tool, data warehouse, email marketing tool, A/B testing tool, ..) accounting for 56 point to point integrations. Each source <> destination connection takes an average of 24 hours to do initial API integration and more beyond to maintain, so hours can add up. (https://segment.com/academy/choosing-stack/how-stacks-evolve...)

We're also buffering data behind the scenes with Centrifuge when partner APIs have a downtime event to get your data deliverability as high as possible :) (https://segment.com/blog/introducing-centrifuge/)

Either way — this feedback is super helpful, and if you'd like to chat more I'm ilya@ segment.com. Thanks!

> For new startups (<2 years, <$5M raised)

Basically this definition works for venture-backed startups, who will fail or get big funding in 2 years. For those of us taking a slower, lower-risk path, two years is a pretty short cutoff for "startup" status.

If the entry-level, non-startup pricing were more modest, this wouldn't be much of an issue. But reading the comments about pricing here, it seems like the non-startup pricing is very high.

One of the characteristics I appreciate and look for most about software vendors is clear reasonable pricing. Temporary discounts, negotiated rates, applications, and the like push me away from products and make it much harder to sell what I want up the management chain.

We did the same thing. $30,000 every 6 months or so I believe. Absolutely insane.

Took 1 day to rewrite.

How did you rack up that kind of price on Segment? Boatloads of anonymous web traffic?

I don’t recall the exact quote. We do have a ton of traffic (more so back then as it was a freemium product then). Not sure if it included their warehouse as well.

Our dev team at the time said rewriting it internally was impossible, so we actually paid it for 6 months. After changing the team, this was on the top of the list. It was trivial to switch.

Now, all the services that we used were stable so the benefit of easily adding and removing services was not something we needed.

Rightscale suffered similar fate when their rates kept going up. At some point it became cheaper to rewrite it ourselves.

> Took 1 day to rewrite.

did you consider making it Open Source?

Lambda + S3?

Yes, I completely agree. It just came across as greedy at the end of the day. Especially given their billing model relies on MTU, when 80% of most websites traffic are fly in fly out visitors.

It's my favourite technology, but unaffordable for pretty much any company not burning some VCs money.

Same thing happened at a company I work at. We got a quote back from Segment, the price was ludicrous, so we rolled our own in a week or so. Increased our Azure bill a bit (500/mo), but still soooooo much cheaper than Segment.

This thread suggests Segment is anchored to their initial value prop for developers here[1]. To this crowd they seem to be more or less a one-way JS wrapper for websites that can later be replaced with a simple "logs to S3" instrumentation.

From my perspective as a "business ops" person, they graduated past this years ago and have unlocked a ton of value with identity normalization and server-side Track events, particular in moving data around between core product and the services marketing + sales use to acquire and engage customers.

Segment is the one place I can manually push an identity update or Track event and know it will be broadcast back to all the other endpoints, including a warehouse. In addition, the services I use also push their user activity back to Segment as a destination, which then automatically broadcasts it back out to everything else.

Asking product engineers to do all this work every time a new need pops up (a) literally never happens, and (b) if it does happen, drives them crazy because they hate working on that stuff and they'll never touch it again once v1 ships.

[1] https://news.ycombinator.com/item?id=4912076

I've used Segment at two companies in a row (in both cases they were implemented at early stages before I joined). It's a good product with broken pricing. They charge insane amounts of money for add-ons and seem to get exponentially more expensive as you grow. My recent renegotiation with them was the most painful experience with a vendor in my 12 year career.

I want to like Segment, I really do. But until they change their attitude I strongly recommend against using them. Especially for startups, because it's not easy to get off once you deeply integrate Segment into your stack.

Had a very similar experience at a previous company as well. At the end of the renewal meeting, we left with no indication our issues would get solved and a quote for a 4X price increase. We churned.

Thanks for your feedback, "painful" is not what I like to hear. If you're willing, will you send me an email ilya @segment.com, would love to get details.

If you run a business, I really don't understand the appeal of sharing a crucial part of your data (for web based applications / services) with a third-party (Google, Amazon etc. or maybe even Segments in the future) that may at some point compete with you ... especially when you can just parse your own web logs ...

Yes, it’s possible to have your engineering team build its own version of Mixpanel. And it may actually cost you less than it cost Mixpanel to build Mixpanel because it’s more focused on your needs.

But you know who has already built Mixpanel? Mixpanel. And they’ll let your business use if for a couple hundred dollars a month. Starting today, not six months from now when your version would be finished.

Better still, if you use Mixpanel’s Mixpanel instead of the one you built yourself, you can have your dev team develop your product during all that time they would have otherwise been building somebody else’s product for internal use.

Build versus buy. It’s a thing for a reason.

> Starting today, not six months from now when your version would be finished.

The raw data is with you. You don't need to wait for 6 months but launch a version with limited features and can continue developing it slowly, until it achieves whatever feature parity you are aiming for.


I'm starting to think that most of these analytics services are more trouble than their worth if you have a good engineering team. A few metrics API's + timescale DB + Grafana goes a long way. If you have a differentiated business you're probably measuring something unique. And if you're measuring something unique then you're better off building your own analytics.

Also, realistically most businesses are built on top of databases. So the most important stuff is in the databases anyway. So might as well just put metrics in SQL too.

Self-hosting your analytics is probably the better way to go. There's Piwik (name has changed to Matomo at some point) and I noticed Countly also has an open-source version your can host yourself. There's a lot of functionality in analytics software and you should not throw it all away and roll your own simplistic thing when you can just take a quality open-source product and host it yourself.

But I agree that applications have unique things about them and for those, people need to implement their own metrics to measure whatever interests them. In that case, the current best choices for time-series database are InfluxDB and Prometheus. They provide very simple APIs and tools for collecting and transmitting metrics, and if you prefer to store aggregations of your metrics instead of the many individual data points, you can use the very simple but powerful StatsD api/client/library/package.

Shameless plug: My company HostedMetrics (https://HostedMetrics.com) offers hosted and fully managed InfluxDB and Prometheus. It includes everything needed for a turnkey metrics platform that is ready to accept your metrics the moment you create an account: StatsD for data transmission, Grafana for dashboards, and the alerting functionalities of InfluxDB/Prometheus.

I am working on rolling out everything on k8s for this reason. I know the conventional wisdom of startups is to just pay for services and move on, but I found that we just keep running into walls of either expense or the need to customize stuff. By the time we learn how to use someone's proprietary setup and get it all integrated, we could have just spun up a container in our cluster to handle it. GKE and a couple lines of kubectl and helm are pretty powerful these days.

The value for pure tech players might be low -as in, you could hack something that does 90%+ of what segment does for you.

But for every other biz that happens to be on the internet, ie e-commerce, not having to rely on a tech team for implementing various services has some value. If you consider that some websites can only be updated every few months or so, having a third party manage that for you is a life-saver.

I've seen / helped too many marketing folks try to get some services (read pixels) implemented and it's always been a Shih-tzuh :)

One product worth checking out might be simpleanalytics.io. its a small company that doesn't track your data. And its cheap. Not entirely sure of your use case but ive seen a good amount of people saying screw the data collectors Ill roll my own and it seems some of what you might be looking for is available from simple (I have no affiliation, just appreciate alternatives to the huge tech companies)


Why are you saying grandparent's comment isn't relevant and is spam?

The product recommendation seems like an alternative to Segments offering (albiet simpler), and the commenter claims no affiliation.

Is there any examples of startups that have failed due to their use of Google Analytics and google competing with them?

Data volume-based pricing is a bit out of control across most analytics/data products. My company has been moving away from volume-based pricing to a feature/value-based model.

The big challenge is aligning data scale to company size/price.

I suspect Segment is trying to solve for this very problem with their price increases but at the expense of smaller business. The 'startup' program is a great way for them to keep the platform accessible to new businesses. I applaud them for that.


Shameless plug: My company Indicative (https://www.indicative.com), a customer/behavioral analytics platform that integrates with Segment. We offer 1 billion events/user actions a month for free for Segment users, so that all companies, especially startups can start using data in bringing their products to market and without having to worry about data limits.

Your product looks great, is related to OP, and your post clearly indicates you are advertising. Good old Hacker News from circa 5-6 years ago would applaud you. Today??

Shame on all those who voted you down.

:) thanks.

This is an email I just sent internally:


Yeah....We need to have a conversation and push back on them-

IMHO, Segments pricing starts breaking down as company's scale because of a couple factors, that generally allow people to push back.

Some of these reasons (that I know of...and I'm sure you know better), are:

1. Google tag manager is free

2. Many services, like Marketo, don't allow Segment to use their universal javascript widget...so you end up using Segment as a tag manager to push that provider's proprietary widget...reducing the value of Segment. In fact, if you look at the "www." side of our website, we could directly instrument 2 sdk's and remove segment (so net +1 javascript library) and dump segment all together. This would also significantly reduce the unique people count...how they bill!

3. We'll soon have a full-time front engineer focused on the website...So it's not very painful to instrument a new sdk, or add a shared cookie between sdk's.

4. There are options that are really good at moving log files (e.g., fluentd.org) for server side logging. It's how the big boys move log files :)

5. According to google analytics...we're at XX,XXX across webapp + www, so not sure why they are saying YYY,YYY

I realize there's value in Segment, but we can push back on pricing increases.

One last thing, allowing us access to our historic pricing, with no additional functionality, doesn't seem like a big give...especially as pricing usually reduces as you scale. So I am very skeptical of this "new pricing" line of reasoning they are trying to pull!

Looking forward to talking to them ~N

I too tried their product, loved it, but couldn't use it because of the insane pricing -- despite me having a fairly small website and just wanting to get better insights to help start making a profit. I quickly exceeded their free limits (within two weeks) and the jump up to $200 per month was just ludicrous. I was also contacted by a very friendly sales/support person but they just couldn't see how spending $200 per month just wouldn't be feasible for me.

Add me to the list of former Segment users. A pricing change made them want to bump our bill from $199 to $10999.

I would explain to them that this was insane. They’d grandfather me in for a quarter. But every quarter I’d get new urgent emails from a new account rep who thought they were going to make a huge score.

We rebuilt event pipes (really not that hard) in house and used stitchdata for pulling from external sources. DIY isn’t without work, but use a hosted Kafka and it’s really not a ton of work. And gives you a ton of flexibility.

I get their niche but I don’t trust their pricing/business further than I can throw them.

Can someone explain to me what Segment is and what it does? Even the website doesn't make it clear. What is it beyond analytics? Can someone clue me into why this is awesome or necessary? Am I dumb for not understanding this or a bad marketer?

its integrations. you drop one javascript snippet in there, and instrument the app with their sdk.

Want google analytics? add it in their UI and its there. Mixpanel? Click a button. Trying a new ad network? Click a button.

Basically, all these SaaS companies want some bits of your data and have some custom integrations.. instead of needing an engineer to implement it your marketing team can edit a UI. And you do it once, so you don't hit as many 'oops we forgot/implemented wrong conversion tracking for Random Service X'

None of it is that hard to do, but there's probably better things your engineer should be doing. Also, as the engineer whose done it, on the tenth marketing platform integration the engineering team will hate the marketing team.

The real value IMHO is the server-side identification and custom Track events, not the JS tracker.

Being able to instrument marketing AND app events that move data around, and then arbitrarily use webhooks and third-party integrations to further move data around on-demand (e.g., via Zapier or customer.io) or in batch (e.g., nightly updates from Looker) is magical for business teams.

I say this as a member of a 6-person venture-funded startup. Segment has helped us massively leverage a small headcount.

This seems like a one-time problem during setup of someone's infrastructure. Is the one-time cost of doing these integrations really worth the monthly cost of Segment? There must be some other value they provide?

>> tenth marketing platform integration the engineering team will hate the marketing team

> one-time problem during setup of someone's infrastructure

As the marketer regularly making these requests, it's by no means a one-time problem, and I can attest to being hated by our tech team (for some reason they don't appreciate me hiding random JS code in our CMS to save on these requests either...).

GTM would solve many of the problems for us as we store all of our customer data in CRM anyway, but for other orgs that don't have a massive CRM policy, having an essentially outsourced data warehouse might be valuable.

I think one of the big ones is that, if setup right, you should be able to change tracking tools and not change your frontend code. They may be banking on the setup cost to be potentially expensive (engineers time) but the buyers are not in engineering (so value the engineers time highly).

You are describing google tag manager, not segment.

sure, segment is just a fancy tag manager. It's similar enough that they specifically have a page calling out the differences: https://segment.com/docs/guides/general/what-is-the-differen...

which amounts to, the data goes through them first, so "Because we store the data, we also have the ability to replay your historical data into new tools and give you acccess to your raw data in a SQL data warehouse."

Ah damn, signed up for segment one week ago or so. Some you win, some you lose I guess.

Echoing other commenters, pricing plans are quite steep for the more modest businesses out there who might have a pretty large reach (Monthly users) but nowhere near the budget to shell out big $$.

Tis a shame cause as someone with a foot in tech and the other one in marketing, I know there's a lof value for the real world businesses out there (outside of the tech bubble) but having to spend what I guess is a hefty amount of cash every month just for the pipes is going to be a hard sell for folks which might be already struggling to see value in investing in technology.

It's okay if you signed up one week ago. Email ilya @segment.com with the details and we'll get something good going!

Will do, cheers.

Been following you guys for a while, cool to see you're expanding -esp here in good auld Dublin and having enough momentum to afford such a move.

Was thinking about using segment here but their pricing is insane.

Many product analytics companies have absolutely crazy pricing. Simply not worth it for the odd experiment, measuring MAU and a few other key metrics.

I think we are likely to setup track.domain.com/event endpoint dumping events into Redshift given that we will use SQL to report anyway.

Segment is amazing. The price point is ludicrous.

I remember a few years ago I was working for a company that wanted to build a Segment clone, just a clone without any different functionality, with hopes they would raise money. Even when I asked how a feature should be implemented, instead of showing me some designs, all they would do was open up Segment and show me the feature... Suffice to say, the project failed and I wasn't surprised nor sad.

Im not surprised though - the cost is prohibitive. And even if they provide a free tool for 2 years, what kind of bill do you expect.

I believe this market is ripe for disruption. A clone, very differentiated by focus would certainly eat away and bring more aligned pricing. Right now, Segment are literally the only folks who do what they do. They do it well, but its not affordable to bootstrapped startups.

We created a Segment account over a year ago, but didn't start using the service because (a) our startup wasn't quite ready, or had the traffic to collect meaningful data, and (b) couldn't afford the pricing at the time.

I wonder if we could transition that old account to this new startup plan, as we are still at early stage (just got into our first accelerator a few months ago) and bootstrapped?

Yep, we can get you in :) Apply at segment.com/startups and send me an email: ilya @segment.com. Thanks!

Segment is great for early-stage startups as it takes care of data pipeline for you so that you don't need to hire data engineers or waste time downloading reports from marketing tools and analyzing data offline using a BI solution or Excel.

Since it doesn't provide any user interface that allows you to get insights from Segment data, you need to use a BI tool in order to be able to make use of this data which also requires you to dedicate resource and spend some time on it.

Early stage startups usually don't have enough time and resource so we have been working on a tool that allows you to integrate Segment data-warehouse and run segmentation, funnel and retention queries. You can think of it as Mixpanel UI for your Segment data. Here is the details: https://rakam.io/integration/segment/

> Since it doesn't provide any user interface that allows you to get insights from Segment data, you need to use a BI tool in order to be able to make use of this data which also requires you to dedicate resource and spend some time on it.

The reason Segment doesn't have a user interface for insights is because it has destinations. There's a slew of BI tools you can choose from the catalog as destinations, including Mixpanel.

If you're using Segment for Mixpanel or other end-to-end analytics providers, that's OK but IMO the real benefit of using Segment is being able to collect all your company data including customer event data, marketing, and CRM into your data-warehouse so that you have the big picture of all your company data.

Our solution basically provides Mixpanel-like interface for your Segment Warehouse Data (See https://segment.com/product/warehouses/) and you have the ability to see all your company data, unlike Mixpanel.

You can always use generic BI solutions to analyze the data collected by Segment but the product data has different semantics and Mixpanel-like solutions help the product teams to be able to ask more ad-hoc questions without any need of help from the data team.

The form won't accept 0 in the total funding raised input. MacOS/Chrome

Hahah, woops we're on it!

clicked on the link to read this, apparently my pi hole has the domain blocked

Same. Seems like they've employed bots to upvote.

I really loved Segment when they launched and recommended them repeatedly to anyone who would listen. I can no longer recommend them because of price and instead suggest people hack something together themselves or use Google Tag Manager. It's just not worth getting your company locked in even if they might give you a startup rate right now.

It was also really frustrating to us doing work in the political and advocacy sector that the their EU GDPR Data Processing Agreement explicitly forbids using their product for any of the things our EU clients would need like measuring petition signatures or donations to causes -- since those might be information about "political opinions" which their terms prohibit. This particular GDPR interpretation is somewhat unusual, we've negotiated workable DPAs with almost everyone else we've wanted to do business with but Segment was unmovable.

This is awesome. Segment has always been a tool that speeds up our experimentation dramatically but the cost has been tough to swallow. I’m grateful your team has built this new program!

I'm having trouble understanding what the use-case for Segment is.

Mind you: I work in a pretty unrelated field and almost never deal with the frontend/web stack.

What would I use it for?

A typical web analytics pipeline involves collecting web data (events like visits, clicks, sign-ups, logins, purchases, email opens, email clicks, website chat sessions, etc.), collating that data and loading the collated data into a data warehouse - we may think of it as the first-mile ETL :)

The web analytics SaaS market is diverse and fragmented and each of those event types (mentioned earlier) is usually recorded via a different SaaS tool. This is where Segment comes in. It provides a unified interface for collection and collation of web data across a wide variety of web analytics tools - Google Ads, Intercom, Mailchimp, HubSpot, Zendesk, etc. and makes this process extremely easy.

This landing page (especially the code/data example) is the best example on why we need something like GDPR. "Look how easy it is, I'll just send your personal identifiable data everywhere at once!!"

They need to define "Startup". What about an established business with a new 'startup'-like product?

We define startups as new companies with <2 years since founding and <$5M in funding. We have soft limits, and we'd love to hear from you either way.

The form assumes you are <2yr/<$5m. So should I email you with details of our company. We are already a customer.

Oh Nice! I’ve been waiting for that.

Back when I was in school there's this guy who sold meth. First hit was free.

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