It's also not even completely altruistic. The free accounts on Cloudflare serve as an incredibly diverse source of first party signal intelligence, which directly impacts the quality of the offering they're able to provide to paid accounts.

The fact that there's at least nominal value to Cloudflare in having me as a free customer gives me confidence and reassurance in using their product, as it lends credence to the long term availability of my free tier account and that I won't suddenly be the victim of a shift in their marketing and sales development budget.

I'm not sure what the right term for that type of incentive alignment is, but it reminds of me Geico auto insurance. As a business, their goal is to provide a long term and stable source of float to Berkshire Hathaway[1], not quarterly profits. To do that, they've optimized their business to capture and retain as much market share as reasonably feasible. Their rates are truly reflective of the actuarial expectations of insuring you - there's virtually no profit margin baked in because that could prevent marketshare growth, and you won't get bitten by shady exclusions buried in the fine print that you only find out about when you try to file a claim, and you won't get hit with an unexpectedly huge rate increase at renewal time due to a slightly worse than expected quarter and the need to engineer short term profits. The "gotcha" with Geico's pricing model is that they're using my business as leverage for other purposes (premiums as float to provide leverage to BH).

I do a lot of work in optimizing monetization strategies and developing/optimizing business processes, so I may be an atypical consumer. But I have far greater adoption of services when I understand the value drivers of the relationship. If you're not directly making profit off of me, I'm either a marketing expense in your acquisition process or you're using our relationship to derive profit elsewhere. If I'm a marketing expense I'll try to avoid building reliance on your product into my life because the stability of our relationship is unknown. And if you're deriving profit from our relationship and I don't understand how, I'll default to assuming an adversarial relationship where either "I'm the product" or there's a hidden catch coming in the future. Whereas if I know that you're not directly extracting profit from our relationship, and I understand how you're indirectly driving profit from our relationship, I'm far more likely to trust you and integrate your product into my life.

" If you're not directly making profit off of me, I'm either a marketing expense in your acquisition process or you're using our relationship to derive profit elsewhere"

Or there's the 'conversion option'.

A freemium user at any given time has a probability of conversion, and that probability is an option on future cash-flows.

Surely, the probability is highest somewhere near the start of use, but not right at the start, as most users probably start with freemium anyhow.

With products like Dropbox which are in some cases 'shared' - there's a material network externalization as well, which is a probability that a freemium account will bring in others. This goes beyond referral/word of mouth to the extent that the product is naturally networked.

For this later reason - social media sites are almost always free because the economic potentiality of the network beyond 'you' is always considerably greater than 'just you'.

Also, as you hinted at in your Geico example, is that there economic realities to incumbency and market share. Once something becomes a market leader and the 'de-facto' option, then there can be incredible leverage, especially in networked situations. What is the current alternative to Facebook today? Not much.

Obviously the dynamic changes once market share is established but before that, there's an impetus keep prices down.

Venture Capital is keen on that to the point wherein one might consider that VC money is just really a form of 'economic dumping' , i.e. money getting thrown into markets to sell stuff below market value to obtain market share. It's illegal in some ways!

The 'conversion option' is just a bottom of funnel component of your acquisition strategy, and the costs to support those freemium users is still a component of your marketing spend.

For established marketing channels/user segments, you should be able to identify your response curves and probability of conversion at each stage of your marketing/acquisition funnel, including from freemium to paid user. If freemium conversions is a component of your sales process, you should absolutely be attributing the costs associated with supporting freemium users towards marketing, net any freemium-specific revenue streams such as ads. It's a fairly substantial component of the fully burdened cost per acquisition, and by allocating the expense towards marketing rather than diffusing it into operations, you get a much more actionable view of your true CPA. Being able to do so at a granular level requires putting in effort to integrate your marketing analytics with your operational analytics. But even just looking at it in aggregate is beneficial to understanding acquisition costs.

Network effects complicate that calculation a bit, depending on the nature of the network effect. Freemium Dropbox users sharing files could lead to net sign ups, as you mentioned. But the value from that is incremental user acquisition, and is still solidly a component of your marketing/acquisition costs. Contrasted with your social network example, where there is intrinsic leverage afforded by the size of the network that's independent/distinct from the potential to drive incremental user acquisition.

VC is its own beast. As you said, it's essentially a form of economic dumping. They prioritize growing market share first, and identifying network effects and associated leverage potential after the fact. Only after all that do they really mature their business model, and does it make sense to try to understand what component of network effects is a marketing benefit vs. a product benefit. Uber is a perfect example of this - their original pitch[1] was very solidly as a taxi app, and one that used exclusivity as a selling point no less. Virtually all of their pitched areas of growth were around potential for optimizing personal transit, with only a single mention of branching out their infrastructure to other location based services applications. Now they've distanced themselves as much as possible from their identity as a taxi service and repositioned themselves as a logistics and transportation juggernaut. And the ubiquity, availability, and consistency of their network is a core component that enables that product identity.

[1] https://medium.com/@gc/the-beginning-of-uber-7fb17e544851

You know it, you're a smart dude

> If you're not directly making profit off of me, I'm either a marketing expense in your acquisition process or you're using our relationship to derive profit elsewhere.

That "elsewhere" might be the word-of-mouth on steroids that I talked about in the article.

Send me an email, tell me more about what you do, interested in hearing - daniel@chagency.co.uk

> That "elsewhere" might be the word-of-mouth on steroids that I talked about in the article. I'd argue that still falls under being a marketing expense. Word of mouth on steroids is a fantastic thing, especially so as a strategy for rapidly generating market awareness and validation of your product. But after that initial validation in the market, it's entirely possible to find your business in a position where it no longer get much value from more word of mouth. Particularly if you've shifted from a passive to an active sales org or shifted upmarket to enterprise sales.

Optimizely[1] is a solid example of that. They used to have a free Starter plan, and had priced plans that scaled from hobbyist pricing level through to small/medium business pricing levels. They've shed all of that at this point in time, and now have three unpriced plans listed on their site and an enterprise sales process could rival Oracle or IBM in the level of self-assured salesmanship and abrasively aggressive price discrimination/discovery. Having used them at a previous company (starting with the free plan and moving to a ~$400/month plan), I was pretty shocked when I reached out to them more recently to adopt at a new company. Budgets less than $50k/year aren't worth more than a one-liner "no thank you" from their sales staff, budgets between $50k - $100k may be enough for them to do you a great favor in allowing you to become a customer, and budgets above $100k/year will at least get them on the phone long enough for them to tell you what the price floor is for your industry and company size (with actual usage volume being next to irrelevant). Word of mouth on steroids became irrelevant to them to the point where they completely shed not only their free tier but also all of their legacy small/medium business customers, too. And it didn't actually hurt their business any, and likely helped it in the eyes of the high margin customers they're going after now.

Contrast that with Cloudflare. Offering a free tier has real costs associated with it. But those free accounts also serve a valuable purpose to their product, and the word of mouth benefits are just icing on the cake. A core selling point of their product is the advanced threat detection and security capabilities that are enabled by the sheer size and scope of their network. The loss of their free tier customers would drastically reduce the visibility they had and in turn degrade the offering to their enterprise customers. Their free tier is a fundamental component of their business/product itself, rather than just an acquisition/marketing/pr cost. Which insulates me as a free user from feeling like the rug is going to get pulled out from under me at any moment when their sales strategy shifts.

The optimizely story was also a slap on the face to me. Maybe their business justifies it, but throwing away your paying customers just feels gross to me.

On the bright side, I would never have created an open source alternative[1] without this happening, so ended up with a better solution for virtually free and hopefully helped others in the process...

[1] https://github.com/Alephbet/alephbet

you mentioning Cloudflare out of all the given examples is the proof that they're giving out a lot — hence why I said I love them despite never using their service

"You will get happy customers who will be happy to recommend you, and happy to pay for your service." — and some of them will become friends. we're talking long-lasting friendships because you helped them get off the ground

thx for the comment!

I thinks the way this ends up working in real life is, know how much value you're creating for your users and find ways to profit that keep them in the black. For many folks, being able to build playlists containing almost any don't you can think of is enormously valuable. And the elegance of advertising in this situation is, those who use the service most end up "paying" more for it anyway through ads. When they're getting enough value out of it that the ads become annoying, switching to a paid tier quickly becomes a great bargain.

That's one way. Another way I was thinking of while reading your comment is something along the lines of "bringing so much value for free, that eventually when they're asking for something from you, you'd feel guilty not to help".

Happened to me with an older guy whom I took advice from. He helped me a shitton without asking me. I'd often tell him "how can I help, you're offering so much free stuff to me". Eventually, he asked me to become business partners in a venture — I said yes willingly but thinking about it in retrospective, there was no way I could say no

What that means at scale could be after a product/online person helps you with a shitton of free resources/content, you'd feel guilty not to pay them back by donating to one of their fundraisers — see Wikipedia

They’re free as fuck and you’ve seen the “please donate so that we can keep it free”. I know they’ve got:

- over $80m (million dollars, yes) in donations in 2015 - assets worth around $100m in 2017 - probably even more due to anonymous donations (and if you’re donating a lot you want to be anonymous)

And I still have absolutely no damn problem with them asking for money. Yes, that’s revenue on top of which there are expenses (a huge bill, given how much they’ve got to cover).

They’re transparent, they don’t hide these figures I’ve just posted and that’s totally fine. Surely they must be in top 1000 value sources of the modern world.

I think it’s not shame that gets people to donate to Wikipedia but, if anything, guilt. I’ve helped you so much I’ll guilt you into returning it, at least to a degree. I think this is the next wave. And then everyone will be doing that (because making someone feel guilty is a flaw, somehow), it’ll be the old-school and we’ll find something new.

But hey, in the grand scheme of things, at least we all started being better by providing value

  My main takeaway is: don't squeeze money out of your
  customers by severely limiting the free plan.

Seems to me how well offering a free tier will work depends heavily on what your product is - and whether it provides a logical place to separate the tiers so users are motivated to pay without losing the goodwill the free tier aims to generate.

For example, if you're a SaaS version control system, it's easy: Charge per-user-per-month for private repos and corporate-single-sign-on-system integration.

Boom, no hassle or difficult sign-up for free tier users, practically every company with more than a handful of developers has to pay, companies with 100x as many employees pay 100x more, no need to annoy free-tier users to push them onto the paid tier, no semi-permeable paywall where users can clear cookies or open multiple accounts.

But if you're a social network, or a tool for debugging regular expressions, or a chat app, or a newspaper, or a streaming music service? Much harder to draw such a tidy line.

Please don't use code blocks for quotations. Reading them is terrible on mobile.

But wait, isn't the biggest social network the perfect example? Made it free for everyone, gathered so much attention

But when you want to advertise, you'll have to compete with so many others, hence the bidding.

One could argue the "you're the product when the product is free", true but the equality of opportunity is there. And the cost of entry is 0.

There's a difference between a product with two tiers and a company with two products. I would not say that Google Search was a free tier of Google AdWords, even if every AdWords user is also a search user, as they're different products.

Likewise, there's not a paid Facebook Premium tier or a paid Instagram Pro version.

The closest things to social networks that have a paid tier within the social media product are the likes of Tinder, Strava and perhaps LinkedIn.