I received a helpful reminder last night from Facebook that they’re about to make some major breaking (but, it should be noted, well-publicized) changes to their app requirements.  Specifically, some authentication code that used to work won’t anymore, and all apps will have to be hosted on servers with SSL certificates.

This is going to be interesting, since up until now, if you had web hosting somewhere, you had everything you needed to have, say, a custom Facebook tab on your company page.  Now you’ll need to pay for a cert, which might not even be available if you’re on a really cheap shared server plan.

I imagine that there will be a few phones ringing on the 1st.  But maybe not October 1st.  The fact is, after a platform’s been around for a while, a lot of “experiments” get left by the wayside, basically abandoned, so for many they might not realize there’s a problem for some time.  For others, this will provide a welcome relief in the form of an excuse to kill off an ill-conceived initiative.

For others, though, there may be trouble if they’ve outsourced the work long ago and need some changes made in a hurry.

Apple has a similar system in place for their app store.  If you don’t pay the annual fee, your apps no longer appear in the store, a fact that I know has surprised a number of independent developers who had free apps in place but didn’t feel like they had to remain responsible for them.

And just like with Facebook, I’m sure at this point every single day a company loses their app (but doesn’t necessarily know it) because it was released into the wild by a well meaning contractor who didn’t see fit to renew.

Two takeaways here: first, are your products and services available “forever,” without the need for users to take some form of action, financial or otherwise, to stay on the books? Forever’s a long time.  And on the other side of the fence, are you in control of your technical deployments?  Disaster recovery planning focuses on large scale events, but “I lost that guy’s phone number” is something that should (yet really, really shouldn’t) be on a lot more organizations’ scenario lists.

In one of those weird Facebook discussions, a very important (relative to the rest of Facebook) question came up:

After a few quick Google searches, I had the answer: 44 million donkeys to 20 million grapefruit, so about 2.2 to 1.

Now, those stats were from 1996, because I only spent 4 minutes on the research (it’s not like it was pirates versus ninjas or anything) but it’s worth noting that the internet has gotten to a point where one can find out the ratio of donkeys to grapefruit in under 5 minutes (and I’ve just made it that much quicker with this post.  You’re welcome.)

Of course, that’s market research about 1996 like it actually is 1996.  These days, there are many more tools at our disposal.

My favourite, hands down, for a ballpark estimate of a potential market, is the Facebook Ad builder.  Let’s see the ratio of people who are interested in donkeys versus people who are interested in grapefruit:

There are a lot of things about this research that’s pretty dumb.  Normally I’d be trying to sell something, so my demographics would be a lot tighter, and I’d have a much richer range of keywords.  If I was setting up an ad campaign, I’d have multiple segments based on individual avatars I’d created, and each segment would have a different reach and ultimately different conversion percentages.  And so on.

All that said, those numbers took less than a minute to pull up.  And they’re tied to a clearer course of action towards the next step, which is generating some traffic around a test offer, building a list, and so on.

Someone’s probably done some research on this, and I suspect it’ll vary by market, but there’s got to be a semi-reliable multiplier to figure out, based on the potential Facebook audience, what the potential real life market size would be.  This is where Facebook is going to get more and more valuable as they crank up the data gathering – assuming, of course, that they don’t start to drop relevance in favour of increasing impressions, but if they did the multiplier would just drop.

So that’s market research, 1996 and 2011 style.  Now what’s 2015 going to look like and how can we get there ahead of schedule?

I know, this post has nothing to do with Eisenhower, but I was searching for Like Button pictures and couldn't resist

Fast Company sends word that Facebook users can now tab company pages in their photos in addition to users, and how it’s the next evolution of Like As they put it:

Social media experts have begun to look askance at the Like system. It’s so easy for someone to “Like” something. What does it really signify? If you “Like” a kind of soda, for example, does that actually mean that you’re buying the item? Or just that you like the idea of the item? How do you really know how engaged with the product any particular Liker actually is?

Now think about tags in photos. If you’ve tagged a photo in which you’re wearing that cute pair of jeans, you don’t simply like the idea of the designer. You’ve actually gone out and put down some of your hard-earned cash for their duds. Photo tags, then, become a much stronger signal of engagement.

I don’t disagree that it’s a stronger sign of engagement; after all, it takes more work than just clicking a button, but I don’t think a tag in a photo means you’re necessarily a customer.  It could just be a sign that you’re more obsessive-compulsve than the average Facebook user.  It could also be a sign that you’re pissed at the company, which means Facebook just opened the door to Twitter-style brand management issues on a platform that possibly has more influence (fun fact: 8% of Americans who use the internet are on Twitter, while more than 41% of Americans have Facebook accounts, apparently – different research methodologies there, but the gap is still significant.)

Anyway, the article goes on to suggest some interesting advertising opportunities, which Facebook hasn’t enabled yet, but could be compelling – if I could put an ad in front of people who took the time to tag my company in their photos, that might be a win, whether the tag was positive (and my brand gets built further) or negative (can’t hurt, really.)

We’re talking much smaller numbers than standard media buys though, given the orders-of-magnitude smaller (yet arguably much more targeted) audience, so unless it can be folded into an existing social media campaign where the daily work is already being done, this might not make sense for many organizations.  I’m a little skeptical of mixing ad buys with social media management in general though; if your team is big enough to have separate content and analytics people, then maybe there’s a fit, but I think the person who does your comment replies isn’t necessarily the person who should be buying your ads – consult him or her regarding them, sure, but it’s a left brain/right brain kind of thing.

Hat tip: Guin. Photo by Kevin.

Does a mobile app make redirects worse or better?

Royal Pingdom asks, is the web heading towards redirect hell?  Basically, as link shorteners get joined by large services wanting to track clicks (clicking a link in Google, Facebook, Bing, and soon Twitter actually takes you to an intermediate step on their servers before sending you to the destination, even if the status bar of your browser says it’s a direct link, and then there are the URL shortening/statistic services like bit.ly) we’re going to have more steps on the way to the page you actually want to go to, which as Royal Pingdom notes, adds time to your page loads and creates many more potential points of failure on the way.

Yeah, it’s going to get worse.

When you download mobile apps for services and they include their own web browser, they’ve got the ability to track links not just on their service, but anything you click anywhere.  This is probably too much data for the average analytics worker circa 2010 to deal with, but as analysis engines grow in sophistication, marketers are going to be able to play 6 clicks of separation (that’s twice this week I’ve referenced 6 degrees, weird!) between their site and any other, or any site and any other, as long as the browsing experience is within their app.  There are some decent advertising opportunities to be had in this space even if we don’t know anything about the user, and guess what, we also know everything about the user…

Depending on the implementation, in-browser click tracking in mobile apps might not cause any of the problems from web redirects, or at least not directly.  The problem is that it’ll highlight the need to track web clicks even more to capture this stuff on all platforms.

And mobile apps aside, as services grow, what’s to stop, say, Facebook from un-bit.ly-ing links?  Most of these are going to spread through their network virally, so one pre-fetch could substitute the bit.ly redirect for their own, so the net redirect count stays the same, but the marketing data from whoever made the original link is gone.  Click warfare could be coming to a browser near you…

Of course, that’s not stopping me from implementing my own custom redirect system later this year.  What?  I’m as curious as the next person on what links work best, but I’d rather be the point of failure than have a service close and kill all my link love.

It doesn’t appear to be available online yet, but the latest issue of Inc has a piece on how to get people on Twitter to follow your business.  One of the companies profiled was Tasti D-Lite, which sells, I don’t know, desserts or something.  OK, I do know.  In the course of writing this post I’ve learned nearly everything there is to know about the company, because fact checking is awesome. But that doesn’t matter right now.

What matters is that they spent $10K to upgrade their point-of-sale system to incorporate Twitter and other social networks into their loyalty program.

Based on the article, this was going to be a different kind of post.  See if you can guess why:

“To get points for tweeting, a customer submits his Twitter username and password. Then every time he buys something at a store, he swipes a loyalty card at the register.  Tasti D-Lite’s point-of-sale system automatically logs in to his Twitter account and sends a tweet informing his followers of the purchase.”

Thankfully, it looks (from the description anyway) that they’re using Twitter’s OAuth integration instead of actually collecting usernames and passwords, so I don’t have to write an article called something like “security schmecurity – people will do anything for ice cream.”

Instead, we’re starting to see early examples of physical/virtual integration, and it’s kind of exciting, if not early (a Twitter search today for Tasti D-Lite suggests that well under a dozen people participated in the past week.)

Lately I’m of a mind that Facebook got people online, Twitter got them more comfortable with telling the whole world a little bit about themselves, and FourSquare made it OK to track their every movement.  The interesting part is that people new to these services find themselves mentally blocked for updates, so the ability to automatically say they’ve bought ice cream is actually a potential value-add as opposed to a minor sellout of their individual identities – it’s really no different from “I’m at Starbucks.”

At the moment, of course, the Facebook – Twitter – FourSquare – Tasti D-Lite (et. al.) funnel is pretty narrow.  For it to grow there’s going to have to be large scale investment in point-of-sale systems, and I can only hope that someone will step in and make something that’s open and extensible to make room for the next generation of players instead of yet another exclusivity play.  Credit card terminal makers, are you paying attention?