Every few months, since late 2012, Snapchat has released a new number to the public. This number is touted as an important, defining metric — it is intended as a sort of Snapchat state-of-the-service update, and reported on as such. It is referred to, variously, as total photos "processed," photos "shared," and photos "sent."
This number is always very impressive. That's why Snapchat shares it. In February, the Times reported that "[m]ore than 60 million photos or messages are sent each day through an app called Snapchat." In April, Snapchat, fresh off a $13.5 million round of funding, shared that the service was "moving upward of 150 million photos... on a daily basis." In September, it was 350 million photos "shared." Today, "the site now processes 400 million snaps per day," according to CNET, which got its numbers from a Snapchat representative.
The CNET story does what many stories before it have done: It compares this number with metrics from Instagram or Facebook. The 400 million figure is particularly compelling because it exceeds, allegedly, Facebook's daily photo uploads. But, the post says, "Snapchat's 400 million figure includes both photo and video messages and it likely counts all "snaps" of the same photo sent to different people."
I emailed Snapchat founder Evan Spiegel to clarify. In response he sent this tweet, posted minutes earlier to the Snapchat account:
Snapchat went on to say that photo/video creation per-day is "not a public metric."
So, yes: The numbers we've been receiving and sharing about Snapchat refer specifically to how many images have been received by users, not how many times users hit send. It's still a very large number, and a number that's climbing rapidly, but it's a very different one from a sort of "upload" figure. Four-hundred million sends would be comparable to Facebook's image uploads. Four-hundred million messages received — or views, charitably — is not. One would assume that Facebook's hundreds of millions of photo uploads are viewed more than once, or that Instagram's tens of millions of uploads are viewed perhaps dozens of times apiece.
The claim that 88% of snaps are sent to a single recipient is intended to assuage worries that the 400 million figure is somehow misleading, the implication being that the sent-to-received ratio is fairly even. But if it were as even as the 88% figure implies, then Snapchat would just share it.
This is the context that most stories about Snapchat seem to miss — particularly now that it is getting mainstream attention, after turning down a reported $3 billion offer from Facebook: We know basically nothing about how big it is.
Here are the numbers we now actually know, for sure, about Snapchat:
• Its users receive about 400 million photos a day
• 88% of photos are sent to just one person
• It has received about $73 million in funding to date
Here are some numbers — each more important than the 400 million figure — that we don't know:
• How many people use Snapchat every day, or every month
• How many unique snaps are created each day
• How many people have downloaded the Snapchat app on any platform
• The rate at which any of these numbers have changed
This is not to say that Snapchat isn't popular. It certainly is, by many reasonable, anecdotal definitions! And the fact that Facebook, which presumably did some due diligence, thought the company was worth three Instagrams, should not be ignored (though it should also be understood as an attempt made by a company that desperately wants, and needs, to stay relevant, and that once tried to rip off Snapchat when the service was very young).
But Snapchat is a startup. It's a startup that is courting buyers, or at least humoring them, and that may be in the process of raising an enormous amount of money. It is clearly in Snapchat's interest to share its most impressive numbers, but it is in everyone else's interest to demand to know more, and to take these numbers — these context-free numbers — with a wheelbarrow of salt.
If you step outside the Facebook/Instagram/Twitter bubble for a moment and try to put these numbers in their proper context, you can produce an entirely different impression.
We can look at WhatsApp, for example, which claimed earlier this year to process over 27 billion messages in one day (a record, sure, but one that is now months old, from a company that is growing quickly). It explained, transparently, that this meant "10B+ msgs sent (inbound) and 17B+ msgs received" — a ratio of 1.7 messages received to messages sent on a service that is functionally very similar to Snapchat, at least in the way it handles individual vs. group messaging. (It is not an ephemeral service, and its user habits are likely different from Snapchat's in many ways.)
Using WhatsApp's ratio, Snapchat users are hitting "send" on photos about 235 million times per day. This is, again, a whole lot of tapping! But to further complicate the issue, WhatsApp claims to have just (well, "just") 350 million monthly active users. WhatApp is a texting replacement service that supports many types of media; its users seem to send a very high number of messages, including, the company says, 400 million images per day — as many as Snapchat users receive.
Now, we're getting into fuzzy territory here, but that's where Snapchat has left us:
• Snapchat says its users receive 400 million photos a day.
• WhatsApp, six months ago and still growing rapidly, said its users received a record 17 billion messages in one day. Let's assume that these levels are now normal — six months prior, the record was 11 billion.
• There are, as of October, 350 million monthly active WhatsApp users.
• The monthly active user/daily message received ratio for WhatsApp is about 1 to 49
• Applying that ratio to Snapchat's 400 million photos received number leaves us with a little over 8 million monthly active Snapchat users.
Again, with plenty of caveats: If Snapchat's monthly user/daily message ratio is anything like WhatsApp's, Snapchat may have fewer than 10 million monthly active users. Halve the daily message assumption — because why not? — and the number is still shockingly low, at under 20 million. Halve it yet again, to an average of fewer than 15 snaps received per day, and you're at somewhere between 30 million and 40 million users. But you're left with a different kind of service, with unremarkable engagement.
That might seem like a lot of messages for a small group of people, but survey after survey has suggested that young people send an astounding number of texts per day. In this context, 49 is not a high number. You could arrive at a similarly surprising number by applying any of these texting surveys — a recent survey claimed young people receive about 60 texts per day — which works out to fewer than 7 million people. My gut says Snapchat has more users than this. But that really just highlights the issue: Our best guess at how many users Snapchat has fallen to a vaguely comparable app and my gut.
All of this is to say, Snapchat's sensational rise into the public consciousness came at the perfect time, just as the startup world has begun to disregard, once more, the laws that govern the rest of the world. Call it a bubble or don't: Snapchat is a poster child for the current crop of exciting startups. Its creators and investors know this, and will use it to their advantage. All stories about, or from, the company should be understood in this context.
Update: More estimates rolling in! Techcrunch has created a scale based on estimated snaps sent, starting at 12.8 million users (sending 20 snaps a day) and topping out at 256 million (sending one snap per day). Appsfire's Ouriel Ohayan estimates a cumulative 75m downloads for the Snapchat app, from which active users is anyone's guess.
Update 2: Another interesting thought experiment, wandering deeper into the haze: If we assume that Snapchat's per-user value to Facebook is similar to Instagram's, for which it paid about $30 per installed user in 2012, then we can make some inferences from Facebook's reported $3 billion acquisition offer. Such an estimate would leave us with a hair over 100 million total installs for the Snapchat app. As for what that would translate to in active users, who knows: the only information we have about Snapchat's level of churn is firmly anecdotal.
Also, in both cases the point wasn't to buy existing users, it was to buy growth. Anyway: Just one more stray estimate for the Snapchat size possibility matrix!
Snapchat has 25m active users according to +GlobalWebIndex. Lead markets for reach are UK 9% of internet users who are active, followed by Australia 9%, USA 8% and India 7%. For teens (16-19s) the reach is much higher with 37% of teens in Australia using, 28% in US and 27% in UK.This, according to the company, is based on interviews with "over 170K users across 32 markets 4 times a year," ages 16-65. According to its data, the USA and China have similar active user accounts, at about 7 million each. The rest are spread around the world.
You can see the full graphic, with penetration numbers, here.
Update 4: Snapchat has responded, in a way:
So, let's look at Onavo! Right now the service says Snapchat has about a 20% iPhone market share. According to Comscore earlier this year, there are about 52 million iPhone users in the country. That works out to about 11 million iPhone Snapchat users in August.
Let's assume the same proportion of Android owners use Snapchat (not fair, but it's the best we can do). Comscore says there are about 70 million Android users. That's an extra 14 million Snapchat users in August.
So, based on Snapchat's recommendation — check Onavo — Snapchat had approximately 25 million American users in August.