Why PPC teams are becoming data teams

Like many people, you’re worried about losing your job to AI.

Where do your “old school” PPC skills fit as AI agents take over more of the work?

Relax. It’s not that binary. The focus is shifting toward data and strategy.

From the outside, it looks like media buying is being automated away. But let’s set the record straight: it isn’t. The role is shifting (again).

I’ve been working in PPC for over 15 years, and there’s nothing to be afraid of. The real question is: are you riding the wave or being left behind?

Let’s map the current PPC landscape: ad network automation and, most importantly, where PPC teams create value today — the critical skill sets and team structure required to compete.

The return of the technical PPC team

A decade ago, technical PPC agencies differentiated through developing scripts, handling data at scale, and managing complex structures. Then automation matured. Everybody started leveraging Performance Max or Advantage+ campaigns because they’re much easier to set up and run.

As a result, many teams shifted toward strategy and creative.

With AI, though, it’s easier than ever to produce good-enough creatives or analyze massive datasets and output what looks like a good strategy. Now don’t get me wrong, those outputs won’t be perfect but:

• It’s free (sort of) and fast.
• The quality level isn’t bad at all (not great either).

From a client perspective, this means the average creative-focused or strategy-nerd agency is out of the game. Those teams need skills AI can’t replace.

So rejoice, PPC people: the technical edge is back. It has morphed into something different for sure. But it’s time to bring back the spreadsheet junkies from the 2010s. They’re the right ones to drive PPC again.

Doubting that? Let’s rewind a little bit and look at the necessary skill set.

The PPC edge: From spreadsheet skills to data nerds

What successful PPC agencies now sell is dramatically different than a decade ago. But the same core mindset resurfaced.

Why?

Let’s look at the core performance drivers these days:

• Integrating down-funnel data into strategy.
• Building a data infrastructure to support said strategy.
• Feeding the right signals to ad algorithms.
• Building systems to operate at scale, including creatives.

See the pattern? You can’t prompt your way out of a broken data model. This is where your edge remains and what clients value.

The good news is that automation increases the value of technical literacy. It doesn’t reduce it.

Who do you call to handle technical literacy? The old PPC marketers. The ones who loved manipulating paid search ads using custom Excel macros they built, or managing hundreds of thousands of product feed items. They have the right mindset: they love automation, data, and math — and they love PPC.

Dig deeper: How to build a paid media team in the AI age

So who should be on your team, whether in-house or agency-side? Here are four essential roles. No single person can cover the entire scope — you need a team.

1. Data engineer

This role basically builds and maintains the infrastructure. Although located after the tracking specialist in the data supply chain, it’s the most central role. That’s why it comes first.

We operate in a complex, multi-platform world: think CRM integration with Google Ads. Or merging online and offline datasets to map the customer journey and drive strategy.

Without a complete data model, your strategy becomes a vague gut feeling that often needs a reality check. The role of the data engineer is to lay the foundation to avoid this situation whenever possible.

Conversely, without this role on your team, you’ll perform repetitive manual exports, get inconsistent numbers across teams, and end up with slow decision cycles.

What is the data engineer’s scope?

Building a data infrastructure basically follows an ETL process: extract data, manipulate it, and make it usable in a reporting tool (think Looker Studio, Power BI, or Tableau).

Here are a few tasks that illustrate that overarching goal:

• Build data pipelines from ad platforms, analytics or CRM tools to the data warehouse (to get spend, revenue and other data into the warehouse).
• Structure tables for those sources and “join” (merge) them to answer specific use cases.
• Maintain those datasets and create automated QAs, including refresh schedules.

What skill sets and tools does the data engineer use?

Generally speaking, since we live in a Google-first world, we hear a lot about BigQuery, Google’s data warehousing solution. There are other solutions, such as Microsoft Azure. However, the main skill set you’re looking for is coding — more specifically, SQL and Python.

The goal here is to use those languages to structure tables within the data warehouse (using SQL) and create data pipelines (using Python).

2. Tracking and measurement architect

Some people consider this to be the same role as data engineers. I strongly disagree.

To me, this role’s sole focus is to protect signal quality. It’s the one person who faces very tight deadlines when things go wrong: you can’t afford to lose conversion data for more than a couple of days. And it’s not retroactive: when tracking is down, conversions are lost forever.

Ad platforms’ performance stands on the shoulders of conversion data. If you don’t get enough of those quality events, you’ll be at a serious competitive disadvantage.

You typically notice this when CPAs fluctuate without explanation or when your in-platform data varies drastically from your “source of truth” (GA, CRM and other systems). Tracking and measurement architects stabilize bidding, increase event match quality and get more data into Google Ads.

What is the tracking architect’s scope?

They design data collection mechanisms that are both complete and regulation-compliant (hello, GDPR):

• Align tracking with privacy compliance.
• Design client- and server-side tracking.
• Implement GTM and server containers.
• Co-manage Conversions API integrations with the data engineer.
• Co-ensure deduplication logic with the media buyer.

What skill sets and tools does the tracking architect use?

Although most PPCs have dabbled with Google Tag Manager, very few have actually set up server-side tagging infrastructure. That’s an easy way to distinguish “regular” PPCs from a tracking specialist. However, they should also be comfortable with Consent Mode frameworks, CAPI, and related tools.

Dig deeper: AI tools for PPC, AI search, and social campaigns: What’s worth using now

3. Data analyst

If the data engineer builds the pipes and the tracking architect protects the signal, the data analyst decides what the data means.

It’s the role most impacted by AI. Granted, you can do a lot with AI, but don’t underestimate how impactful a great data analyst is.

The wrong interpretation can waste millions of dollars in a blink of an eye. Fully replacing data analysts with AI would be a gross mistake.

For example, ROAS in Google Ads doesn’t equal contribution margin. Meta Ads CPA doesn’t equal customer lifetime value.

Without a strong data analyst, you risk misinterpreting data and going down the wrong rabbit hole. Think cutting campaigns that look inefficient short-term but drive long-term value. Or reporting different “truths” to marketing and finance — you don’t want that.

What is the data analyst’s scope?

People outside the field think they build Power BI or Looker Studio dashboards. That’s just the tip of the iceberg. Data analysts also:

• Design data models aligned with business KPIs (this step kind of overlaps with data engineers at times).
• Run analysis — think cohort performance, churn rates, profitability, and diminishing returns.
• Challenge platform narratives.

What skill sets and tools does the data analyst use?

I tend to think of data analysts like translators: you can speak another language somewhat fluently, but that doesn’t make you qualified to interpret at scale. Same with data analysts: you may understand numbers to an extent, but you probably still need an analyst.

SQL literacy is often required to query the warehouse directly. Spreadsheet modeling also remains critical for scenario planning. The key skill is statistical reasoning. Understanding sample size, variance, and bias prevents false conclusions.

4. CRO and experimentation lead

Once all that data is clean, available, and analyzed, CROs leverage it to improve the economics of every visitor. Improving conversion rate, lead quality, and the overall customer journey creates a compound effect.

The simple way of proving CROs’ worth is to understand that a landing page that converts at 1.5% instead of 3% means you’ve doubled your CPA. Nobody wants that. And that’s where CROs come in. Instead, you want to scale efficiently, not push more money toward a leaky bucket.

From a PPC standpoint, CROs strengthen both performance (better conversion rate) and signal quality (more conversions), which helps smart bidding.

What is the CRO’s scope?

Contrary to common belief, CRO doesn’t (solely) mean landing page. This role operates across the full funnel:

• Mapping the journey from impression to revenue.
• Identifying online friction points using heat maps and session recordings.
• Structuring testing roadmaps instead of random experiments.
• Collaborating with creative and product teams on offer positioning.

What skill sets and tools does the CRO lead use?

The entry stack I see most often is GA4 and a heatmap tool such as Hotjar. However, it can get much pricier with tools such as ContentSquare. The stack scales depending on the client’s needs and budget.

The skills that matter most are:

• Just like data analysts, a deep understanding of math and statistical reasoning (think pre-calculated sample sizes).
• A structured mindset, clear hypotheses, and business-level success metrics.

Dig deeper: Agentic PPC: What performance marketing could look like in 2030

The modern PPC team looks less like media buyers and more like a hybrid between marketing, data, and product. The advantage goes to teams that structure these capabilities deliberately.

Winning PPC teams are the ones who understand algorithms, but more importantly, the data and economics behind them. If your team masters infrastructure, signal design, analysis, and experimentation, AI becomes leverage. If not, it becomes a liability.