Freemium did not work for Phanfare
blog.phanfare.com
blog.phanfare.com
We lost the ability to effectively use CPC search marketing (google). When we had a free trial, it was easy to see which clicks were worth paying for. But with freemium, the conversion funnel was so long (average of nearly a year before the person needed more storage) that any attempt to optimize price per conversion was hopeless. (Try adjusting the shower temperature with a 12 month delay between knob and temperature change and see how it goes for you. In control theory parlance, this is known as introducing a delay in the feedback loop)
It's very interesting. Anyone used to CPC lives in a world of certainty. That kind of certainty has rarely existed in marketing. The problem with that kind of certainty is that taking out risks eventually takes out returns. If you are selling a dropshipped commodity good, margins drop (cpc bids go up, prices go down) to more or less 0.
Freemium on the other hand, slots conveniently into the old world of TV campaigns & billboards. You have no idea if you did good or not. That kind of risk keeps out the number of competitors needed for 0 margins.
If you decide to go for the freemium model there are several things you should check before taking the plunge:
- are other companies offering what you intend to offer behind the pay-wall for free or not ? - are you going to be able to monetize your free users in some other way ? - what is your expected conversion rate from free users to paying users and can your business survive on that rate (+ a healthy margin for safety) ? - what is the cost of supporting a free user in case the monetization of a free user is less than the income on that user ?
A fairly simple spreadsheet will give you the answers, if you can't determine the conversion rate then do a survey with a pledge. That way you get a ballpark figure, reality should be within 50% of that figure, so to be on the safe side discount it by a factor of 2.
Last but definitely not least, if your 'premium' package offers nothing of value to your user (because for instance multiple free accounts give the equivalent of a paid account) then don't bother trying.
They should have thought this through before going ahead with it.
- To create a large community with network effects that increase the overall value. Here it's hard to charge anything at all, especially if you have any kind of competition. Once you have lots of users, you try to think if any of them might be willing to pay for something.
- To give out a Super Sample. Get free users, when a small percent will convert to paying members. This is Chris Anderson's 95%/5% freemium. Here the line is fine.
The important point about freemium is that it's interacting with a competitive environment. You probably cannot become a major photo/video sharing site without being free, because others are. This is a result of low marginal costs.
Acceptable is anything over 1:2000, 5% is two orders of magnitude above that and I have yet to come across a company that uses the freemium model that has that kind of conversion rate.
Do you have an example ?
I'll look at what I use. Both Weebly & Dropbox (I use the free of both) follow a true freemium model. IE, they both offer a perfectly usable free version. I'm not sure what percentage they do, but 1% - 10% sounds about right to me, especially if you discount those who are really sampling vs those who actually use.
I also use Gmail & Gdocs. These are theoretically freemium too, but at a much higher free/paid ratio.
You're going to see quite a lot in the professional/business software category. My explanation for this is that it is software people pay for. If you can sell your software/service, maybe can probably sell it with a freemium model.
If you can't sell it (eg photo sharing/search/etc.), maybe you just can't sell it. Freemium isn't magic.
The percentage has to relate to your marginal costs in some way. Even 1:2000 can be fine if your marginal costs are zero. You care about conversion rates only inasmuch as you care about absolute numbers.
*I recoiled a little when I read Anderson quoting an actual ratio. I think that it's a bit off. But I do think segmenting businesses by their ratio would be an interesting way of looking at them. A 5% business isn't following the same model as a .01% one.
There are a lot of different subsets of products that you could sell using the freemium model (content, raw infrastructure, access to applications and so on), I figure there are probably ballpark figures for the conversion rates for each of these.
What the big question is to me is what the 'optimal' conversion rate is and what the path to achieving that is.
Another huge factor (and one that the article completely overlooks) is what the competitive landscape looks like. If you're in the business of providing some freemium based item (say free email with a 100MB inbox, more is paid) then you're toast if you get a competitor with deeper pockets that gives away the same product that you are charging for.
Storage and services related to it (and bandwidth) will converge on '0' so you will need a lot of users on the input side of the funnel in order to make it work for the few that will max out your caps and that don't want to game the system by setting up multiple accounts.
They probably decided to compete on a numeric basis with the other 15,000 undifferentiated Store Your Big Files Online sites. Competing on numbers is the same as competing on price, i.e. stupid.
OK, maybe that is a little harsh.
But if you reduce yourself to competing on numbers, any numbers, you'll always be at the mercy of the biggest fool in the industry, who will one-up your numbers. Additionally, the customers you get because your numbers are momentarily the best are disproportionately pathological customers: they have no loyalty to you, they have loyalty to The Best Deal, and when you are no longer The Best Deal they will move to the guy who is.
This is the same problem that plagues webhosting at the low end. $10 at our service buys you 1 GB of hard disk and 1 GB of bandwidth. No, wait, $10 at our server buys you 2GB/1GB! No wait, $8 at our service... and before long people have bid it down to stupidly low numbers (< $4), a million people are on a box, reliability is terrible, and folks are churning madly.
I think the key for selling commodities like this is to figure out how to decommoditize your product. MP3 players were commoditized, then Apple came out with one which was a status symbol. Hosting is largely commoditized, but Slicehost has a scalable product which merits non-commodity pricing.
I would think if you go freemium, you at least have to offer a compelling "free". 1 GB seems useful, 250MB seems useless.
Interesting how that stuff goes, and also good to have a nice, handy list of all the "when not to"s.
The people you need to offer something to are the ones giving you money. All the rest you're supposed to be enticing to become subscribed users.
Shouldn't they have known how many paying customers were below the 1GB threshold and been able to anticipate all that lost revenue? It seems like it should have been obvious from the start that the 1GB free was too high for them. Why not 250mb or 100mb or 50mb or whatever was below where all those high profit paying customers were sitting?
Freemium, as I understand it, is about providing a useful free service but with constraints that are hit by serious users fairly quickly. Useful enough that the product is usable and will entice a user to want to pay to keep using it, but with constraints enough that it won't be useful forever if they keep using it (i.e., they'll hit the limits and have to pay). Seems like for Phanfare, 1GB was way too much space -- users were finding the service useful, but never hitting up against that constraint.
The debate has mostly been around whether or not it is a viable business model, not whether or not it is always the best business model.
People may have no problem paying $10 a month for a service, but when that same service was previously free or at least free at there usage level there is going to be a backlash and if you are in an ultra competitive area like this one that could be a big problem.
So if you needed 5GB it is relatively easy to use UserName01...UserName05 to acheive that.
Each additional 1GB faces a slightly reduced problem :: Is it worth coughing up cash for this service??