This is part 1 of the 3-part series titled Data Activation Is Not The Goal.
Have you noticed that sometimes, a term that has been around for a long time seems novel only because its usage suddenly starts to soar? This is particularly true for technology categories, and one such category is Data Activation.
What's really interesting is that the term "data activation" has been in use for a while – only the context and meaning have changed. Looking at trends for the last 10 years, its usage peaked in 2016.
People in Sri Lanka searching for how to activate the data plans offered by their respective carriers. 📶
That brings us to the first part of this series…
The Commodification of Data Activation
Data Activation is a topic I’ve been covering since early 2020 – after all, activating data is what got me deep into the data space when I was leading growth at a fast-growing SaaS startup.
However, what I didn't anticipate at the time was that the term “data activation” will soon be used to describe a new category of data tooling.
As a growth person, I took it for granted that the act of running data-powered campaigns and experiments was in fact, “data activation” – essentially going beyond analysis and taking action on the available data.
Taking action on data or activating it seemed like a natural next step after running analyses in a product analytics tool and identifying points of friction in our user journey.
The goal was straightforward – to move users down the funnel by getting them to perform the next best action.
And the process of activating data wasn't very complicated either – we delivered in-app guides and emails, triggered based on user behavior (events) and personalized based on the user persona (traits).
For instance, users who didn’t save a workflow (our activation event) would enter an email campaign where the email body contained links to in-app guides that were personalized based on the industry the user belonged to (which they told us during the onboarding survey). These in-app guides or walkthroughs would then get the user to perform the desired actions, leading them to the aha moment.
We were also sending events from the tool powering the in-app guides to our product analytics tool to understand the impact of those walkthroughs on user activation. Doing so enabled us to know how many users completed a guide before saving their first workflow, which further allowed us to improve our in-app guides.
I’m sharing this context because it’s important to highlight that growth people have been “activating” data long before the practice became popular, without any of the tech that’s associated with “data activation” in 2023 – namely, CDP, Reverse ETL, and the latest entrant, Warehouse-native Apps.
It’s also worth mentioning that during my time as a Head of Growth, we were collecting and activating a whole lot of data even without a data warehouse, using very basic homegrown tools.
It worked alright and we were able to get a lot done. But we also ran into a lot of limitations.
Activating data without a data warehouse has its limitations
What has really changed in the last 3 years, thanks to the availability of better data tooling and the adoption of the cloud data warehouse, is that the possibilities of activating data have increased manifold.
But that’s not all.
Today, it’s also a lot easier for growth people to activate data, especially if they can work closely with their data counterparts.
I want to highlight that this evolution isn’t just great for GTM people, it’s equally great for data people – they are now able to work closely with their GTM counterparts to build the most efficient experiences powered by the freshest data, thereby making measurable contributions to business growth.
Does this imply that the practice of “data activation” is making data people think like growth people and vice versa?
I certainly think it is.
I also believe that organizations that are able to harness the power of this confluence between data and growth are the ones that will be able to thrive in a brutal economy like that of 2023.
Now, coming back to the limitations I was referring to…
Combining event data with object data
Without modern tooling, it was extremely tedious for us to combine event data (behavioral data from our web app) with object data from our engagement and support tools. I’ve written about this before in case you’d like to dig deeper but essentially, we had to spend a lot of time extracting and wrangling data from various tools to understand how our campaigns impacted user behavior.
A cloud data warehouse not only makes it easier to measure the impact of campaigns (assuming campaign data flows back into the warehouse) but also makes it much faster to iterate on experiments.
Moreover, with data from all relevant sources available in the warehouse, working together becomes frictionless for both data and growth people.
Syncing data between third-party tools
We had an internal service that would track behavioral data from our web app and sync that data to the tools in our growth stack. However, sending data between third-party tools was harder than we thought (even though our product was an iPaaS solution and we thought we’d easily be able to drink our own champagne by building connectors for the tools in our growth stack).
I wanted to set up a sync between Zendesk and Customer.io to make sure that users with an open support ticket are removed from any ongoing email campaigns. It seemed like a no-brainer that users shouldn’t be nudged via email to try something in the app while they waited for an issue to be resolved.
However, implementing this seemingly simple use case was so complicated that we didn’t end up implementing it at all. Syncing data between the tools wasn’t the biggest challenge – it was syncing the data between the correct entities.
And that brings me to the biggest limitation we faced in the absence of a data warehouse…
Working with account-level data
As with any standard B2B SaaS, our product entities had a many-to-many relationship between them – a workspace had multiple users and one user could create or be added to multiple workspaces.
But that’s not all.
Many of our agency partners were managing over a dozen workspaces and were asking for “agency” accounts that would allow them to centralize billing across workspaces.
And there’s more.
We were in the process of introducing the concept of “teams” into the hierarchy because customers didn’t want to give every user access to the entire workspace – a reasonable ask.
It’s certainly not a good practice to aggregate user activity across all the workspaces a user belongs to – actions in Workspace A shouldn’t impact their experience with Workspace B. Moreover, a user should be able to specify the communication they’d like to receive at the workspace level (instead of making assumptions about what’s valuable for them).
Fast forward to 2023, a cloud data warehouse combined with visual audience building and reverse ETL capabilities makes working with account-level data so much more straightforward.
P.S. We also offered private cloud instances and were working on an embeddable version of the product as well – the thought of working with all these different entities in the absence of a data warehouse makes my head spin.
So many ways to activate data
I’d like to conclude by iterating that there’s no right way to implement “data activation” and one doesn’t need a specific set of tools to “activate” data. Just like one doesn’t need a specific set of tools to “analyze” data.
For early-stage startups with limited resources, sending event-based emails by syncing data directly to the ESP – and doing it well by focusing on good copy and timely prompts – is enough “data activation”.
On the other hand, for growth-stage organizations, using a data warehouse in conjunction with a CDP – either packaged or composable – is key.
The next part of this series is called The Stairway To Data Activation where I offer a concrete definition of Data Activation and explain how it is one of the activities that lead to growth – give it a read and let me know what you think.
The section below contains some additional material on Data Activation.
Data activation tooling has come a long way since my growth practitioner days. For the last two years, I have been sharing an annual post on how the various categories and tools are evolving and converging.
A lot has changed since the previous update from April 2022 but what’s most notable is that pretty much all leading CDP and Product Analytics vendors have built stronger integrations with cloud data warehouses, particularly with Snowflake.
If you’re curious, have a look at my article covering Data Activation since 2021 (including a 2023 update), and let me know what you think is coming in 2024.
I’d once predicted that Reverse ETL tools will eventually begin to compete with another category of tooling that started as PLG CRM, evolved to PLS (Product-led Sales), and is now being referred to as different things by different vendors.
It seems that this particular category hasn’t found a strong footing in the GTM/growth stack (one of the early companies is also shutting down), and we’re yet to see what happens as the lines between data/growth/GTM stacks continue to blur.
However, Tim Geisenheimer, the co-founder of a PLS product, recently shared that many of their customers integrate their product with Reverse ETL tools and explained why that is the case. Whatever PLS tools end up calling themselves, looks like they’re solving a very specific problem for enterprises with a hybrid (product-led and sales-led) GTM motion.
Talking about Reverse ETL, as someone who has worked with both the leading vendors in this category, I always believed that Reverse ETL is a feature that ultimately benefits GTM teams, and it wouldn’t thrive as a data integration solution aimed at data engineers.
However, I wasn’t quite sure how Reverse ETL products would evolve to offer a more complete solution bringing data and growth teams closer. Now I do.
As part of BrandJams (paid collaboration between databeats and brands), I recently interviewed Boris Jabes, the CEO of Census, where we discussed what led his team to narrow down the problem space to Reverse ETL (before the term existed) and how this piece of technology has quickly evolved into a feature that’s become table stakes for all of Martech.