[lubos]

MixPanel has very generous free plan and many free users. Sales people were hired to try to convert free users into paying customers. Plan failed, sales people shown the door.

This is why companies need to think really hard about freemium model. In case of MixPanel, many of their free users keep adjusting so they send just enough data to stay under free plan. They are probably not going to be persuaded by sales people to pay.

[bane]

One of the things lots of people (including founders) don't understand is that when they're converting from early stage VC subsidized business models to "we need to actually be a real business and make money", maintaining freemium means that the subsidies move from the VCs to paid customers, and paying customers hate to cover freeloaders.

[jaredtking]

This. We use Mixpanel's free plan and were recently hit up by their sales team. It's unfortunate because Mixpanel is an amazing product that we would be happy to pay for, but it's just too expensive relative to the value we get out of it.

[jay_kyburz]

I spent a day hooking up Mixpanel. Went over quota quickly and discovered that, instead of just dropping data when you go over quota, they lock your account and blackmail you to becoming a paying customer.

I can't remember my conversation with them, but it was clear they were a bunch of a-holes I didn't want to do business with.

I tell everybody I can to avoid them.

[api]

Sounds like broken freemium: margins too low to sustain, price too much of a leap.

[Swizec]

As one of those free MixPanel users - their non-free plan is just too expensive.

I eventually stopped using MixPanel altogether. In a choice between "only get a small subset of data" and "get all the data, but super super expensively", I decided I just don't care enough and opted for no data. Or rather, the combination of gAnalytics and a few others is Good Enough (tm).

[aaronbrethorst]

I had the same experience, and I have to say that I'm torn in that I can see both sides. MixPanel doesn't want to screw around with $10 or $20/month plans, but I have no interest in paying $150/month.

They want bigger customers than me, which is fine, but that also means that if and when I turn into a bigger customer, I might already be locked into a different solution.

[anatari]

We were one of these potential bigger customers willing to fork over thousands per month, but the unit cost was still too expensive for us. We tried to negotiate a better bulk rate, but their team was unresponsive and so we eventually went with another service that was more responsive to our needs. According to the article they only have ~3k paying customers, which means they only care about giant top tier customers willing to pay top dollar. Seems like a mistake to me, but what I only know from the potential client perspective.

[azinman2]

Either you really aren't getting value, or you don't have many users of your own. $150/mo is much cheaper than building and maintaining something equivalent. $150/mo is nothing compared to even payroll taxes, let alone payroll itself.

[aaronbrethorst]

    or you don't have many users of your own. 

right. yet, hopefully.

[jusben1369]

I hope folks realize that's ok. Mixpanel made a conscious decision to start at $150 and understands that for certain people theirs no ROI at that price point. I mean Apple could sell a $300 iPhone but they'll leave that to Android.

[jfoster]

For the reason you suggest (lock-in during growth), it seems to me that they are being foolish by not having a startup tier. Tens of thousands of $10/mo or $20/mo customers adds up to quite significant revenue.

[robryan]

Maybe this is where they are heading? The heavy sales team route requires larger customers for it to be worth their time converting them to paid users.

They might be looking to cut back on sales and try a more automated approach to convert people to smaller plans.

[aaronbrethorst]

It can also add a significant customer support burden, which is likely why they haven't done it.

[claar]

This was exactly the path our startup took as well.

[Impossible]

I had a really bad experience where a Mixpanel salesman aggressively cold emailed me on my personal email (I'm guessing he got it through Linkedin?) and continued to do so even though I told him not to contact me multiple times and that I had no ability to make or influence purchasing decisions for analytics software. I understand hustling, but at the time it came off as really desperate.

[throwaway2048]

well it looks like his job was on the line

[jusben1369]

I always kind of chuckle when people assume a salesperson tracked down their personal email vs the much more likely scenario that you used it during a trial signup and just forget. At my company a customer's assistant demanded to know why we charged her bosses personal CC for a subscription because "he never uses that for work ever!" Smh...

[lyime]

If you are doing 25,000 of fewer events, you are probably not making much money.

[buro9]

That's not true.

When I looked at MixPanel we were already at more than 8 million page views per month and wanting to track more events than that.

But as we wanted to try MixPanel, the best way was to start by sampling a small number of users and activities.

As our trial of the software progressed we found we got way more value out of sampling a smaller set that we could make sense of rather than measuring everything.

For us, MixPanel became one more tool in a wider toolset. We didn't see the value in jumping up to the paid level as we didn't feel it would replace all of the other tools.

[iyn]

Can you elaborate on the tools in a 'wider toolset'? I'm always eager to learn about new tools/services.

[lubos]

You actually get 200,000 free events if you put their badge on your website. The next tier is 500,000 events for $150/mo.

[alextgordon]

It's not a population of 25k events, but a sample size of 25k.

News organisations pay a lot of money for polls of only 1000-2000 people. 25k data points is quite sufficient in most cases.

[staunch]

Isn't Dropbox having the same problem?