When a business can see activity but can’t explain growth

If you have several agencies, a stack of pixels, and no idea which ad actually made you money, you are paying for activity, not growth.

More spend won’t fix that. Another agency won’t either. Fix the data first, then build one view that shows what is working. After that, every dollar you spend gets a little smarter instead of a little louder.

If that sounds like your account, book a 30-minute call. I’ll tell you where I would start.

More tracking can mean knowing less

A DTC brand I came in to help had hired several freelancers and agencies over time. Each one added its own pixels and its own CAPI setup. None of it was feeding the business anything useful: most of it sent partial data, and some sent nothing at all. There was no naming convention and no account structure. Nobody could say what was working, and the team couldn’t understand why they were losing money. It was a ship without navigation.

The odd part is that the extra tracking made things worse. Several sources giving partial answers are harder to act on than one source you trust, because each partial answer gives someone a reason to believe their channel is the one working.

The first job was fixing the data. The second was a dashboard that turned that data into a story the team could act on. Within a couple of months, the account had turned around.

How it shows up

When a system can’t learn, the symptoms look like separate problems. They usually share one cause.

• Tracking exists but isn’t trusted. Conversions and landing-page events fire, but nobody has checked them. Platform reporting ends up carrying the decisions.
• Campaigns have no shared structure. No consistent segmentation, exclusions, retargeting lanes, or funnel stages, so nothing can be compared across channels.
• Optimization happens when someone has time. Bids, budgets, placements, audiences, creative, and product economics all need regular attention. Without a cadence, they drift.
• The landing page doesn’t match the ad. The offer is buried, the CTA is unclear, and the page could belong to any brand.
• Lifecycle is an afterthought. Email, SMS, referrals, and repeat purchase sit unused while acquisition keeps paying for new customers.

Patch these one at a time and you get more activity and the same confusion.

The order I fix it in

1. Make measurement trustworthy. Validate tag management, pixels, conversion events, and landing-page firing. Fix server-side or alternative measurement where it’s needed. Set naming and tracking conventions before anything depends on them.
2. Give media a structure you can read. Rebuild campaigns around audiences, exclusions, retargeting, and funnel stages. In e-commerce, put margin, category, bestseller status, and inventory into the media decisions.
3. Make tests comparable. Name hooks, CTAs, audiences, concepts, and angles the same way across channels, so a test in one place teaches you something everywhere.
4. Fix the step between the click and the sale. Simpler language, the offer higher on the page, a clear CTA repeated, and a page that matches what the ad promised.
5. Let lifecycle carry the gains forward. Behavioral email and SMS, better retargeting, referral, and repeat-purchase programs turn one sale into the next.
6. Read every stage from one decision layer. Pull and clean cross-channel data on a schedule, and use it to decide what to scale, change, or stop.

What changes

The same sequence is documented in a separate engagement, the luxury DTC case study on this site. There, return on spend and conversion rate rose together. Average blended ROAS across the engagement was ~5x, up from ~1.2x during a weak period. Landing-page conversion rose from ~8% to 33.5% after the rebuild. Those figures are approximate, from engagement reporting, and they belong to the whole system, not to any one campaign.

In both cases, the bigger change was how decisions got made. They stopped depending on whichever platform dashboard someone opened first.

Five questions for your own account

• When a platform and your analytics disagree, do you know which one to believe?
• Could you compare your last ten creative tests using the same names and definitions?
• Does the first screen of your landing page say what the ad said?
• Does a first purchase reliably lead to a second one, or to a referral?
• When you scale, change, or stop something, what evidence is that based on?

If more than two are hard to answer, spending more won’t help yet. Book a 30-minute call and we’ll find the first thing to fix.

That engagement, with its scope and evidence limits, is in the luxury DTC case study.