October 03, 2013

Mobile: A Developing Problem

We'll address the graph in a moment.

A key tenant of the Merchandise Forensics framework is that item reductions, especially via new items, leads to an erosion in customer productivity. And an erosion in customer productivity leads to fewer new customers, and fewer retained customers. This leaves us with a weaker customer file for next year, which drives down the volume per item/sku next year, which causes the merchant to "hunker down" and offer fewer items/skus - which causes fewer customers to purchase. The death spiral is underway!

I observed this in 16 of 20 recent Merchandise Forensics projects, FYI. This is a problem. It is especially a problem with catalogers trying to conserve cash - investing cash in paper instead of investing cash in the stuff customers purchase - merchandise.

But I digress.

So far, I don't like what I'm seeing, when it comes to Mobile. Love Mobile, don't love what it forces us to do, from a merchandising standpoint.

Let's go old-school ... back to catalogs. Go talk to an Inventory Manager, and you will hear a common lament:
  • "If we don't feature the item in a catalog, the item does not sell well online."
In other words, the shift from catalogs to e-commerce caused a shift in the distribution of item selling characteristics.

In old-school cataloging, you saw something like this:
  • 50 Great Items.
  • 100 Good Items.
  • 200 Low Performing Items.
Catalog marketing thrived on having a large number of good performing items - these items filled the catalog, making paper reps happy, making printers happy, driving down cost per thousand pages circulated, increasing new customer acquisition circulation (making co-ops and list folks happy).

Everybody wins when there's a bounty of good performing items.

E-commerce changed the dynamics of item performance. If you don't have a catalog to support advertising for good performing items, you end up with a different distribution:
  • 60 Great Items.
  • 65 Good Items.
  • 225 Low Performing Items.
Google changed the equation - moving focus to the items that Google's algorithm favored (search). Google essentially "hid" the good performing items (page 6 of results). Email changed the equation - moving customer focus to the items that were given the biggest discounts and promotions. E-commerce eroded the number of "good" items, but rewarded "brands" with an increase in great items.

Mobile has many advantages over e-commerce, and myriad advantages over cataloging.

But mobile has one, enormous, whopping drawback.

In Mobile (as it exists today), you can only feature so many items. Given the display size, you are really, really stuck with what you can and cannot display.

Oh, I know, the pundits will argue against this (in catalog, they'll point to Catalog Spree - but if that was so great, why do so few people use it and why did they have to partner with Google), but be honest. On a 4" screen, there's only so much you can feature. The message, the story, it all must be simple. You can feature 100 product hyperlinks on a home page or landing page. You can feature 700 products in a 124 page catalog. You cannot do either (well) in Mobile (and maybe you shouldn't do this in e-commerce, either).

When what you feature is limited, then what the customer purchases is limited as well.

Based on the data I'm analyzing (highly preliminary), the distribution of items is going to continue to change. Here's what Mobile promises us:
  • 80 Great Items
  • 20 Good Items.
  • 250 Low Performing Items.
In other words, we're seeing a "hollowing out" of the meaty middle, where old-school catalogers generated a ton of profit (allowing tens of thousands to catalogers to be profitable). Instead, we're moving to a world with a handful of best sellers, and a whole bunch of low-performing items. If we have limited space to display our merchandise (Mobile screen), we'll optimize that space with best sellers - it's a Darwinian process. This means we'll erode the "good" items - they either graduate up, or they drop down.

This brings me to the image at the top of the blog post.

When we measure the relationship between skus and $/visit (multiply conversion rate by AOV to yield $/visit), we see a clear, linear relationship. If all things are equal:
  • Offer 100,000 skus, yield $6.50 per visit.
  • Offer 200,000 skus, yield $8.00 per visit.
Now, the cost of having the 100,000 additional skus required to bump $/visit from $6.50 to $8.00 might be prohibitive. But if you figure out how to deal with the inventory (i.e. you don't own it), then the relationship is highly profitable.

But this is e-commerce, where a hundred thousand skus can be viewed via a series of drill-downs and landing pages and links, via considerable home page and landing page real estate.

How does this change in a Mobile environment?

Well, it changes everything, as the pundits say.

You're only going to feature a limited number of items. It's all you can do.

Mobile will likely limit what the customer can see. Look at your own phone - how many apps do you honestly use on a continuous basis? 15? 3?

As a consequence, Mobile will drive an effective reduction of skus - you'll still offer 100,000 skus, but your customer will perceive your brand as having 3,000 skus, or 300 skus, or more realistically just 30 skus.

Sku reductions lead to lower customer productivity, and fewer new customers.

In order to make Mobile work, we'll have no choice but to increase traffic.

Not everybody can increase traffic.

So in an arms race to capture a finite amount of Mobile traffic, we'll discount and promote ourselves into oblivion.

This leaves us with a handful of winning brands with enormous traffic (just like Amazon in e-commerce, but think something even bigger than Amazon in the future, or think Amazon becoming 40% of e-commerce instead of 20% - 25% today), with everybody else fighting for table scraps.

The handful of winning brands will be able to personalize, using sophisticated and expensive technology, to show off numerous skus - each customer will see a different assortment. This has been promised for two decades - but the technology is getting closer each day (experts will tell you it is already there, and it probably is), and Mobile will demand use of the technology to drive up conversion rates.

Remember, I write this stuff because I want you to think. 

Don't dismiss this argument with nonsense like "we're different, customers love our brand". Think about the argument. Think about the ramifications of the argument. Think about merchandise. I'm telling you what I'm observing, across a diverse set of clients.

Think.

Discuss.