Like Online Ordering, CDP Is Now Table Stakes Infrastructure. Both Should be Much Cheaper.

If the pandemic made one thing abundantly clear it’s that retailers (especially restaurants) were significantly underinvested in digitization.

And it’s no surprise: surveys have showed that retailers have been laggards in IT spend for years. Not just laggards, but with the exception of perhaps one industry, THE laggard.

The problem is that a 2, 3, or even 4% difference in IT investment matters.

A lot.

Particularly in low margin environments like retail.

Because IT is often very scalable, generating outsized returns for relatively little investment.

And as time goes on, things only get worse if you haven’t been keeping pace.

For example, if you bought a new car today we bet the car would come with power windows.

Honestly: can you even buy a new car with manual windows anymore?

(Turns out there are five models of car that will allow you to buy manual windows as of 2020 :/ )

But given that there are hundreds of car models on the market, we wouldn’t bet against the trend of automatic windows becoming, well, automatic.

Clearly the cost of automatic windows were more expensive than manual windows when they came out, but over time the cost has come down. Today you’re paying a small premium to receive a power window in your car, but it makes driving so much more convenient that it’s rarely a second thought.

If you do have a manual window and it breaks in a few years, it’s going to be more costly to repair the damn thing than it will to just replace it with an automatic window. Why? Nobody will have the parts, repair technicians won’t be trained in fixing them, and the manual window will be a relic from a bygone era.

Technology often goes the same way, only faster.

Bet you have a smartphone, don’t ya?

Personal computer?

See: things that seemed totally bleeding edge only a decade ago are now commonplace.

And so goes the same with CDP.

CDP is just a glorified database; we’re not putting that thing on a pedestal.

But it’s the foundational piece of infrastructure that retailers need to get their customer data together.

Especially retailers that 1) aren’t digital-only, or 2) can’t afford a top four consulting firm to collate their customer data.

The interesting discussion from here is two-fold.

First, who owns the CDP budget?

Second, what should it really cost?

It’s our view that CDP is actually an IT expense.

Read that again.

Today most folks sell CDP to marketing departments. And yes, while marketing departments are a very common user of CDP, CDP is fundamentally an infrastructure expense.

For example, who owns the budget for online ordering today?

Yes, there are clearly marketing use cases, but because of the integrations and data flow requirements it’s very much an IT budget item.

CDP is the same, although we’d contend that CDP better frees up data for finance and operations use cases, which typically get little love with online ordering.

Second, CDP is going to be really cheap in a few years.

Like super cheap.

Because the cost of CDP comes (ostensibly) from three places:

1. Data systems integrations
2. Data storage
3. Query cost

Data integrations are a one-time cost. Unless you’re dealing with true ecomm systems, there’s unfortunately no such thing as a clean API. Human intervention is required to make sense of things, and time costs money.

Data storage costs are cheap and only falling. You can store terabytes of data for dollars a month, especially if you’re reasonably competent at file storage optimizations.

Lastly are querying costs. Complex segmentation can be expensive if data architecture isn’t well designed. ETL vs ELT – it can get pedantic (but dogmatic).

However, well designed systems shouldn’t cost much to operate… think on the order of $1 per location per month for a well designed system.

If you’re scratching your head and wondering why CDPs are most expensive than that the answer is simple…

Sales cycles!

Retailers take forever to make decisions.

It’s why investors hate companies that target retailers: it’s impossible to get the returns math to work.

Customer acquisition cost and lifetime value never pencil. And despite there being tens of billions in potential TAM on paper, it never happens because retailers by and large won’t pay for what things are worth (hint: that’s why vendors make their pound of flesh in the payment processing stream or by renting a retailer’s customers back to them).

We’d love to see a public business model where CDP costs like $1/store/mo just to observe if sales cycles change.

We’d bet not, because retailers don’t self-educate, but we do expect that CDP costs will shift and the money people are paying for CDP today will be gobbled by something that’s much less infrastructure-y and much more ROI-tangible.

By the way, how is online ordering not already close to free? It should follow the same logic: commoditized offering, tenuous ROI, very little cost to operate.

Yet somehow the costs seem to increase year after year.

Maybe the sheer fact that retailers spend a pittance on solutions has prevented competitive vendors from entering the market and kept prices high?

We know larger restaurants are definitely grumbling about Olo’s costs… perhaps this plays to the advantage of companies like PAR that will now offer a compelling alternative.