March 01, 2016

Omnichannel Costs


Pundits misunderstand the cost structure of channels. There is a reason certain merchandise is sold online, and certain merchandise is sold in stores.
  • Retail has horrific fixed costs.
  • E-commerce has horrific variable costs.
The omnichannel thesis, then, is highly dependent upon one key customer behavior ... without this behavior, the thesis crumbles under a mix of fixed and variable costs.
  • Existing customers must spend more, and there must be more new customers who are willing to shop each channel, in order for the omnichannel thesis to deliver incremental profit.
To date, neither has happened. Existing customers tend to spend less over time (ask Macy's, the omnichannel flag bearer), and new customers (Millennials) choose merchandise/experiences over channel experiences.

When neither happens, all you do is transfer sales from one channel to another, and that doesn't work for anybody.

So, today, we sit here with increased expenses (variable) and decreasing store sales (as customers move online) causing stores to de-leverage (fixed costs).

This will cause stores to close (ask Macy's and/or Wal-Mart, for instance).

When stores close, fixed costs disappear, but debt remains.

When stores close, the sales do not all shift online and/or to other stores. You are lucky to recoup 30% of your sales, on average. The fixed costs are gone, the debt stays there, and the sales that are left are largely variable in nature (if they go to other stores, you cover fixed costs ... if they go online, they are tagged with added variable costs).

Without a dramatic gain in merchandise productivity, the math simply doesn't yield growth ... the math can yield increased profit, in some instances, depending upon how poorly the lower quarter of stores perform and how sales shift when the poorly performing stores close.

Only the very best retailers know the math, and understand the right path into the future.

It is highly unlikely that anybody who doesn't work at a retail brand truly understands the math and the dynamics, unless they've written tens of thousands of lines of code to thoroughly understand the dynamics.