Five Ways Serious Performance Marketers Should Use Their Data Today

The mild (was it?) panic about Google sunsetting third-party cookies has now faded, but it hasn’t completely gone away. Now that we know we’ll retain access to our flawed, incomplete, disconnected data, at least for now, we’re anxious instead about extracting meaningful insights from it and acting on them.
It’s almost like some people were hoping they’d lose third-party cookies so they could use their absence as a scapegoat for poor insights.

Not you, of course.

Yes, our data sometimes feels inherently impossible to unify and understand, and that’s a huge pain in the butt. But I’m here to say: I think people sometimes get distracted by that problem and miss the actions they could be taking even without solving it.

The Status Quo of Performance Marketing Data

If you’re still relying solely on data inside platforms like Meta or Google (aka second-party data), the same warning applies to you.

The current state hurts. Don’t do nothing.

Attribution: The Classic Example

You check your dashboards and see a 300% ROAS on each of your main channels: Facebook, Google, and TikTok.

You’ve spent $5,000 on each, so that means you’re pulling in $45,000 in sales…right? Maybe. But more often than not, the real number is lower because each platform takes credit for some of the same conversions.

Beyond Challenges with Attribution

Depending solely on the limited data each ad platform provides means your data is:    

• Fragmented:
  Data lives across platforms, tools, and teams, which makes it hard to stitch together in meaningful ways.
• Short-term focused:
  Most platforms only show and track performance in narrow windows.
• Biased:
  Each channel grades its own homework with their own definitions of success, so results rarely align.
• Incomplete:
  Privacy updates to iOS 14 and regulations like GDPR have left gaps in what platforms can track and report, including cross-device behavior, app activity, and offline conversions.
• Opaque:
  Black box algorithms hide the path-to-purchase and your true value chain.

All of these things together mean it’s hard to assign credit to a single channel or campaign.

But they also mean that the most important parts of your job are harder, too: uncovering insights, deciding which ones to act on, and setting strategy.

You can solve these problems.

What Every Performance Marketer Should Act On Now

1. Control and Own the Data You Can

Every brand has some first-party data: your CRM, the data you collect directly through interactions on your websites, apps, email campaigns, etc. Finding even small ways you can make use of and prioritize that data in your decision-making can lead on a path away from black-box-plaform-dependence.

The degree to which you invest in technology that can help you unify, organize and act on your data can vary.

Here’s a crawl/walk/run/sprint:

Crawl: make the data you already have earn its keep (15-20% ROAS lift)

When: Now

How: Start with the customer records that sit in your CRM and the purchase events your pixel or server-side Conversions API is already capturing.

Export a list of recent buyers or high-value leads, upload it to Meta, Google, or TikTok as a Customer Audience, then let each platform build its own look-alikes. Pass your purchase events back to the same platforms so they bid on real order value, not clicks.

This single move sharpens targeting, lifts match rates, and gives you a taste of first-party power without new infrastructure.

Example Stack:

• CRM (HubSpot, Klaviyo, Salesforce, etc.)
• A pixel or CAPI set-up
• Occasional CSV or native connector to keep the list fresh.

Estimated Lift:

• 15-20% ROAS

Walk: AI gives you better insights faster, and makes those insights actionable (20-40% ROAS lift)

When: As soon as you have baseline performance benchmarks.

How: Once the manual list uploads are paying off, feed those same CRM and pixel signals into Pixis. Add a live product feed and a handful of business fields (margin band, predicted LTV, for example) so Pixis can start adjusting bids by the hour, rotating creative, spinning new audiences, and logging every test result back into the next campaign.

You move from “upload and hope” to continuous, AI-driven optimisation across bidding, audience discovery, and creative.

Example Stack:

• Bid & budget optimization (Pixis)
• CRM
• Ad platforms
• Product feed from Shopify, BigCommerce, or Feedonomics.

Estimated Lift:

• 40% ROAS (Footwear brand) to 70% ROAS (Betabrand)

Run (a marathon): close the loop with richer context and governance (40-60% ROAS lift)

When: You are ready for enterprise-grade orchestration and have the operational and legal bandwidth to support investment in more, richer, first-party data.

How: Build a robust data quality, cleanliness and completeness apparatus to provide your systems of action (like Pixis) with even more granular, accurate data to act on.

Example Stack:

• Data warehouse (e.g. Snowflake, BigQuery)
• Event capture, Identity Resolution (e.g. Segment)
• ID Graph (e.g. LiveRamp)
• Reverse ETL (e.g. Hightouch, Segment)
• Data quality (e.g. Monte Carlo)
• Bid & budget optimization (Pixis)
• MMM / Incrementality Platform (e.g. Recast, Measured)
• Consent Management (e.g. OneTrust)

Sprint (an ultra-marathon): use AI to optimize every granular action and insight in real time (60-80% ROAS lift)

When: Your enterprise-grade data stack is still bottlenecked by your ability to extract meaningful insights quickly and act on them in real time.

How: Move past a single source of truth and into the land of distributed truth. Train a Large Language Model (LLM) specifically for performance marketers, then give it access to all your various systems containing any relevant data. Use it to close any remaining gaps in context in your data model, or between tools.

Example Stack, instead of or in addition to the list in “Run”:

• Real-time cross-channel analysis and insights (Pixis)

No matter how hard you lean in, first party data can help you make better decisions, and provide richer context to targeting/bidding algorithms and AI in your ad platforms.

2. Let AI Make Sense of the Mess

I mentioned this above in that “sprint” section. Just wanted to show you what it looks like.

Automation and rules-based tools require structured data and clearly written rules that reflect that structure to work well. That’s what drove the market-wide obsession with a “single source of truth”.

The idea was that if you could get all the data clean, with one big, neat data structure, and you could make it accurate, complete, and up-to-date, then all your automation rules would work well every time and you would improve the efficiency and effectiveness of every system in your martech stack.

AI (Large Language Models especially) are particularly good at making sense of unstructured data.

They also benefit from not being overwhelmed by the number of levers you could pull to improve results on any given day. Bid strategies, budgets, creative variants, audiences, margins, channels, product availability, messaging tests - AI has no trouble keeping all of these variables in “mind” at the same time.

That’s why Pixis (who build the AI systems at the heart of Social Hustle and all Stellar agencies) are turning their entire focus to building a cross-channel AI for performance marketers.

It’s awesome. It looks like this:

3. Be Dynamic with Your Creative 

First-party data doesn’t just improve reporting; it also changes how you think about creative, too.

Lots of brand marketing teams get tempted to orient themselves around building an interesting, beautiful, desirable brand. Sometimes that’s wise!

Other times, that objective leads them down a path of prioritizing subjective opinions about the quality of their creative. Even when they’re ‘data-driven’, I’ve seen teams become obsessed with optimizing for tier-B metrics like likes, clicks and shares. Those things might get them a shout-out or an industry award.

Conversions, demo-requests, purchases: that’s what gets you promoted.

You own that data. You know what customers bought things. You can feed that to your ad platform, and let it optimize for that, rather than trying to exert so much control over “likability” of your brand that you unintentionally kneecap your own campaigns.

So try it: pair your first party data with ad formats designed to convert, like Dynamic Product Ads (DPAs) and shopping campaigns that automatically update based on real user behavior, intent, and interest.

4. Know Your Numbers

This is trite to say, but I’ve seen it so many times that it bears repeating. Sometimes you really just don’t need more data to make better decisions.

If you don’t truly understand your unit economics, you could be quietly losing money.

Imagine this scenario: 

• Product selling price: $1,000
• Cost to product: $800
• Customer acquisition cost: $400

ROAS is 250%! Nice work.

But when you add up your actual costs, you’re spending $1,200 to sell a $1,000 product. Not good.

I run into so many marketers who don’t understand what the business targets are. If you own an ad budget, you should be able to speak to your target and performance against that target for each of these metrics at any time (just for a start, and in addition to ROAS):

• gross margin
• customer acquisition cost (CAC)
• lifetime value (LTV)

You can’t just optimize for what looks good on a dashboard because most dashboards only show you lead measures (not the most important lag metrics). If you want your data to subsequently drive better results, you have to go deeper and align every strategic decision with your business numbers, not just what the dashboards tell you.

5. Adjust Your Mindset

When you realize what’s already possible to act on with your data (messy as it may be), this should happen on its own. But it does pay to stay conscious of your mindset.

Consider it a self-limiting belief to think that your messy data is what is keeping you from having better insights and/or acting on them. Write it on a post-it note by your desk if you have to.

But stop believing you can’t do better just because your data is a mess.

Own Your Data and Your Results 

We’ve dodged the third-party apocalypse (for now), but the warning sirens are still flashing: you cannot rely solely on the third-party data that lives inside the major ad platforms. 

It’s fragmented, short-term focused, biased and incomplete.

And most of all, it’s holding you back from making smarter investments, better decisions, and driving tangible outcomes for your business.

The good news is that the solution is clear: own your data by taking action on what you can today. Luckily, that list of things grows longer all the time. In a dynamic landscape where everything seems to be slipping away, your data is the foundation you can rely on.

Ready to build stronger strategies and drive real outcomes with more dynamic data? Request a demo today.