In the previous article in this series we discussed the types of CDPs that should be considered based upon the type of business you are. In this last article we’ll discuss the evolution of CDPs and where CDPs are going so that we might give potential users a glimpse into how they should holistically evaluate a CDP solution.
Let’s start off with a polarizing declarative then back it up with facts.
CPDs command no pricing power because they’re a glorified database.
The moat to building a CDP is narrow (and shallow), and as such there’s a lot of competition. Customers rightfully pay little for a CDP because the ROI is tenuous.
A CDP is not directly attributable to making or saving money.
Read that statement again.
Only marketing automation tools (with measurement) can determine if your marketing is making money, and analytics tools that identify opportunities for cost reductions can determine if you’re saving money.
Neither of these require a CDP.
They do require access to cleanly formatted data, but that could just as well exist in an Excel spreadsheet.
CPDs have realized this, and that’s why you’ve seen consolidation and feature bundling over the past half decade.
Note: entire cycles of the business economy are just phases of bundling and unbundling. Typically, once solutions become commoditized they must get bundled with additional value adds to command any value. Then some new breakout feature/product emerges in the category and the market moves to unbundle.
All that to say that CDP is very rationally in its bundling phase as CDP software has now been around for over a decade and is, without argument, commoditized.
So what are CDPs doing to bundle and drive ARPU?
This deserves a graphic.
Below are what we see as the six steps in reaching marketing nirvana. You can see that CDP is in the middle, as it’s a piece of infrastructure that aggregates and houses all the customer touch points needed to achieve maximal ROI… but by itself it offers very little ROI.
Step 1
Customer data – and metadata – are queried from the CDP by a marketing orchestration tool. In an ideal world this tool combs for customer patterns to automatically segment users and build campaign flows that maximize ROI.
For example, Bob buys mousetraps quarterly. Sometimes he buys them online, other times he goes in-store. He opens marketing emails most often on Friday mornings and opens texts most often Tuesday nights.
Based upon Bob’s behavioral and purchasing patterns a marketing plan is put together specifically tailored to Bob to achieve the optimal ROI without – hopefully – being flagged as spam by Bob.
Step 2
The segments and sequences from the prior step are injected into a marketing automation tool. Acronyms get fuzzy depending on whom you talk with, but marketing automation can also go by CRM (customer relationship management) or ESP (email service provider).
Step 3
The responses to the marketing automation are captured by the CDP and the customer records are updated. In other words, Bob just got a text message and it’s the job of the CDP to see if Bob opened his text message and how he responded. Did he come into a store to buy a mousetrap? Buy one online? Flag the text as spam?
Step 4
The responses from the marketing automation are measured in an analytics or measurement tool. This tool once again queries the CDP for updated data so that it has the latest data for measurement.
Step 5
The measurement tool can and often does work in conjunction with the orchestration tool to optimize the results. Turns out Bob loved the text and we grabbed an incremental purchase from him. We can only know it was incremental – and not cannibalizing – by applying data science over longitudinal views of both Bob and similar customers that were held out as controls.
Score.
Step 6
The learnings from the measurement is pushed back into the CDP – and the orchestration tool – and marketing goes back in to optimize further.
If you think about all the machinations, this ends up being a soup of solutions:
Segment.
Iterable.
Amplitude.
Klayvio.
You probably recognize some of those names as e-commerce martech unicorns.
But who can afford all of them? Especially as they’re becoming commoditized.
Bring in the bundling.
Twilio acquires Segment so it can offer a CDP behind its marketing automation tools (which also come via acquisition).
mParticle, a CDP, adds marketing automation and insights with an acquisition of Indicative.
Amplitude, a marketing measurement tool (again, fundamentally for e-commerce) builds and launches their own CDP.
And don’t forget that every large, legacy software company offers some flavor of a CDP: ADP, Oracle, Salesforce, et al..
This is why merchants need to be cognizant of the overall trend in the martech universe.
Do you select multiple point solutions or align with a partner that’s collapsing the stack via bundling? If you choose a solution that’s not bundling, are you prepared to be left behind lest you source more parts of the stack elsewhere?
Is the bundled solution good enough for your use cases or do you truly need the extra 10% of value of each, independent, best-of-breed solutions provider? Can you afford it?
And more importantly, do the providers have the ability to collect all the data you’re going to need to successfully identify, execute, and measure your marketing?
If you have a brick and mortar presence, the answer is almost always no (even with large, Tier 1 merchants [ >$1B of payments volume], there are ways to optimize data collection).
Because the only way to successfully bundle the martech stack for offline merchants is to understand the payments infrastructure so deeply that you’re given access to nearly every card swipe on the planet.
This lets you identify the markets where there’s marginal utility, uncover all your unknown guests, and then deterministically measure the success of your efforts by measuring important metrics – like guest retention, share of wallet, etc. – against your local market.
Oh yea, that sounds like Aben.
But not everyone cares to understand ROI, because they don’t want to be measured on their performance. Nothing like guaranteed income, for little work, and no concerns of reprisals for underperformance, amiright?
For everyone else that demands fiduciary responsibility, there are only so many options; actually only one.
Add comment