Wednesday, November 14, 2012

I'm dreaming: supermarket shopper

One of the most frustrating experiences are to scramble through the superstore to find an item that's hiding. more often than not, nobody's there to help you. So you go through your shopping list and then you end up scrambling to find what you had not found yet. Another issue is to find the place where your shopping cart will be cheaper. I'm guessing the latter would not be in the benefit of the grocery store chains...

In any case, here's my ask:
1- Create a shopping list based on recipes I "add", my past shopping habits, how long typically it takes supplies to wear out etc. There are more than a few apps in this space (Link)

2- Tell me where my shopping list would be cheaper (Link)

3- When I go there, guide me through the isles and tell me exactly where each item is. You know where I am (Link). Now mapping all the grocery items is not trivial. It could be done by the store, but as said, they are unlikely to cooperate. You could tag each item with RFID transmitter. Unsure that will work and how many consumer phones exactly have RFID receivers? What you could do is crowd sourcing: Let each customer tag the location and price of each item and collect all the data. Drive sufficient benefits to the crowd, and they'll map it for you.

In return for this service, I'm more than happy to share my historic shopping cart with the grocery store chains, happy to see complete shopping cart coupons etc.

OK, whose on? ;)

Saturday, November 10, 2012

Mobile revenues and user behavior

I'm lumping few topics in this post, bear with me ;)
First, I came by this interesting article on "Fat fingers' based on GoldSpot media. 38% of total clicks on static banners are "fat finger" effect! Holy smokes!!
What's interesting about it is the form by which they measure accidental clicks: "engagement with post-click content lasted less than 2 seconds".

Triggers a few thoughts: mobile ads can lead to many places, including the app store and/or another mobile website. How many of those load in 2 seconds? close to none. Therefore, I assume that the measurement is really after the page completed loading.
In that case, the user had much more time to look at pieces of the page as it was loading. According to Compuware, over 77% of "top companies" mobile pages takes longer than 5 seconds to load (link). The average response time for a site, again according to Compuware, is 8.1 seconds (link). That means that PostMedia users really have something between 2-10 seconds to view content.

The other point to keep in mind is that we're talking here mobile advertising, which is a part of overall mobile revenues. To give some context, according to emarketer, advertisers will spend a relatively small amount of money on ads on phones and tablets this year — $2.6 billion, less than 2 percent of the amount they will spend over all. Yet that is more than triple what they spent in 2010 (link).
Compared to $2.6Bn in advertising revenue, according to eBay CEO, eBay and PayPal mobile to each transact $10 billion in volume this year (link).

I still believe mobile advertising is far more effective compared to desktop advertising, but I commend MediaPost for coming forward with the reality and attempt at measuring the Fat Fingers phenomena.

But it leads to the core issue (IMO). We don't really understand what users are doing in mobile. We need to dig deeper on how users engage mobile web and native apps and be able to better measure the funnel of all the complex scenarios in terms of experience performance and conversion. Some new innovation can be seen by ClickTale's mobile heat map, but there's much more to do in this space.
Of particular interest (to me) is the cross-channel revenue generated, originated by users on mobile. I believe that revenue portion dwarfs that of direct mobile revenue, and thus mobile plays a much bigger part of the overall company revenue.
Who can do it? let's see. Google? mostly yes. They know everything you do on Android (YES THEY DO), and given Chrome is the world's most popular browser (=shopping machine), they can tell. Apple also has a lot of control as many iOS users will use Safari in some form, probably on mac. Same story.
So it can be done, and it's important.