Building vs. Buying A Data Platform: Insights from Wild Earth's former CMO

A conversation where we dig into the prospect of building vs. buying software. There are many nuances to consider when considering building tech in-house, which we discuss, along with whether the pros outweigh the cons.

KEY
STATS

"If you're not clear what it takes to build your own software, try it buying first."

Steve Simitzis

Former CMO

“We were just connecting random tools and hoping to make a narrative out of that.”

"I wish I had found SourceMedium sooner."

“Any great eComm data tool must be both CEO and team friendly, and that's SourceMedium.”

Build or Buy? A Scaling Brand’s Million-Dollar Question

One of the great debates in “startup land” is whether you (an early-stage company) should “build or buy” a piece of software.

For Steve Simitzis, former CMO of Wild Earth, this is a question he’s racked his brain to answer (and probably lost sleep over, too).

Having been with the company for over four years, Steve has been the caretaker of all things data at Wild Earth, from joining when there was no access to reliable attribution and retention data to creating running queries and pulling Shopify reports.

And ultimately deciding whether it was Wild Earth’s (a now 9-figure brand) time and money to build their own data solution.

We sat down with Steve to talk about his journey to Wild Earth, the lack of access to quality data, the challenges of balancing the ROI of building a data company within a company against the backdrop of hiring and outsourcing decisions, and more.

Piecemealing spreadsheets together

Wild Earth needed the initial technical infrastructure to know what core levers were driving the company forward.

“When I first got here, there was absolutely nothing from a tech and data perspective,” admits Steve.

“I remember we had this Geckoboard on a flat-screen TV in the office showcasing random sales graphs daily. The CEO would walk past it, notice that numbers were in the green, and shout, ‘It’s a good sales day!’ The next day, he’d see numbers in the red and tell us that we should be concerned about sales being down.”

But Steve knew there was more to the story.

As he dug into the details, he discovered that much of the sales data reflected customer subscriptions, creating a false perception and narrative of how the marketing team performed with sales.

So, Steve set out to hack together a DIY way of pulling data from different sources to understand better how the company was performing and what changes they could make to fuel future growth.

“My first ‘innovation’ at Wild Earth was to take Shopify data, send it into BigQuery, and categorize by first-time customers, new subscriber customers, and recurring customers.”

Looking back, Steve believes that that simple decision changed Wild Earth’s approach to business in many respects, all for the better.

Going forward, it set a standard for how each department used data to make decisions, from finance and operations to brand marketing performance and ad buying.

Evolving past Shopify reports and roadblocks along the way

But pulling Shopify reports only brought Steve and Wild Earth so far.

“I call it (Shopify reporting) ‘vibes-based’ retention analysis,” Steve grinned.

So, they evolved from a Shopify-only approach. First, they began adding tools like Looker Studio, integrating subscription data from Recharge, running their own SQL queries, and more.

“We were just connecting random tools and hoping to make a narrative out of that.”

After the initial setup, led by Steve and Wild Earth’s COO and CFO, their DIY approach proved unsustainable for the company’s growing demands.

“I was running around like a crazy person meeting with team members, running queries, maintaining the website, working with our head of growth, managing email, helping the CX team, all while keeping up with my day-to-day,” recalls Steve.

So, they hired a data engineer to alleviate the pressure valves.

And it worked, to an extent. Then came the roadblocks.

The challenges of finding accurate data with a growing team

The company found itself in a meteoric rise in the pet space, on the verge of fundraising, but needed more clarity around where and how to resolve its data woes.

A few challenges they faced:

1. Confusion on how to define attribution, as each tool had different black-box methods
2. Stitching together several horizontal, disjointed data tools that were often too expensive to justify the setup and maintenance
3. Pulling and slicing data from various sales channels to neatly create cohorts for retention analysis

Even with a data engineer, they couldn't do what they needed.

“We realized that the data was becoming a hugely important part of our business. Everyone wanted answers to questions about data, and we only had one team member to field the barrage of requests,” noted Steve.

Steve and his team encountered a critical set of requirements that all cross-functional organizations face:

1. The finance team handles the most critical data to allow the business to function and help clarify demand planning, reporting to investors, and fundraising purposes.
2. Meanwhile, the marketing team wants everything from attribution and ad conversion to retention analysis, and their needs constantly evolve.

“We’re a dog care brand, not a data company; so when we needed to decide how to allocate resources, we needed to consider whether we should hire another data person or put that budget towards CX, Marketing, or Ops.”

Build vs. buy: a burning question

“All I knew is that we wanted answers to data questions at our fingertips.”

You’re a growing company, actively hiring, and you need precise, accurate data to inform decisions that affect the business’s future.

But at what expense?

Ultimately, Steve decided buying software was the more sensible for growing brands like Wild Earth.

From keeping up with constant schema changes to ensuring their transformation layer stayed up to date with Wild Earth’s evolving business needs, accruing technical debt was the last thing Steve wanted.

“I’m so jealous of anyone who starts an eComm business in 2024 because the tooling available to marketers and data practitioners is night and day just a few years ago.”

And, to Steve, what separates SourceMedium from other competition is the platform’s user experience.

“Any great eComm data tool must be both CEO and team friendly.”

But he’s not simply referring to pretty dashboards and nice aesthetics. It’s about offering detailed filtering and drill-down capabilities through pre-built, well-thought-out transformations, enabling technical and non-technical operators to splice the data in whichever way they need to.

Anyone can build another Tableau or Looker dashboard. But creating a product that enables you, the user, to pull rich, deep insights that actually can move the business forward?

That’s what Steve believes is the great differentiator.

Trial by fire

Steve regrets not discovering SourceMedium sooner, and that feeling underscores the challenges of building custom software.

“The technical debt of building our software has proved quite challenging.”

Building a custom solution sounds ideal for scaling brands that are lean in resources like Wild Earth until the accrued technical debt is too much to bear in the long run.

How can one keep up with the evolving business needs and data variables at each growth inflection point?

Perhaps the answer is that you hire more engineers.

But for Steve, over-indexing on data hires when faced with the problem of allocating that budget to bring on talent across marketing, fulfillment, or growth is a decision he couldn’t justify.

That’s where SourceMedium enters the conversation. It’s become a data command center for Wild Earth, built to scale with them, even as their omnichannel data needs change.

Return to customer stories

Unlock e-commerce growth with turnkey data solutions

Request demo