Why the most “successful” campaigns aren’t driving revenue for brands
Michael Leppan, Co-Founder
The tension between CMOs and CFOs (or agencies and their clients) usually boils down to one thing: what’s being reported vs what’s reflected on the bottom line.
High conversion rates and top line figures look great on dashboards, but when they don’t deliver profit, inevitable questions arise about the effectiveness of the campaign, sometimes cascading to marketing function as a whole.
Unequivocally, it does work. We know this because our entire business model and our success as an agency is built on delivering rapid, profitable, incremental growth for clients. If we couldn’t achieve it, and if we couldn’t prove it, Open Partners wouldn’t be here.
Rather than overexplain the problem, below, I detail how CMOs can address this challenge, outlining the biggest mistakes marketers make when setting up, optimising, and reporting their campaigns and what they can expect to achieve through varying levels of data maturity.
Mistake #1: Optimising against gross revenue instead of net margin
To understand why performance marketing breaks down, you first have to look at how ad platforms operate.
Today’s AI driven marketing platforms are incredibly sophisticated: you assign an objective, and the algorithm optimises delivery to hit that target as efficiently as possible. But as platform algorithms operate as self-contained systems, they naturally evaluate success based on the conversion data returned to them (essentially marking their own homework).
The algorithm isn’t broken; it’s simply executing the instructions it was given. If the machine learning engine lacks the internal financial context of what is truly profitable, it can’t differentiate between a high margin transaction and an unprofitable one. It will take the shortest, most efficient route to deliver the exact goal assigned.
The most common “instruction” we see (and amongst the biggest campaign setup mistakes) is optimising for gross revenue instead of net margin. This overarching instruction essentially forces the algorithm to do two things:
- Target repeat buyers who would have converted anyway, because it’s the path of least resistance
- Chase high order volume over order value
If you are an ecommerce brand, for example, this means the algorithm targets your existing customers, and regardless of what they spend, will give the system a “thumbs up.” The platform will continue to optimise very effectively against its target and where it knows it can deliver impact, but remains blind to whether those sales are delivering true net profit to the business.
Mistake #2: Underutilisation of business and product data
Above, we mentioned “instructions.” By that, we mean data. If you continue to tell the algorithm to optimise against data that looks great but isn’t really meaningful, how can you expect it to deliver profitable performance?
These algorithms need to be trained on rich business and product context in order to be successful. In practice, here’s what proper application of business data can look like:
Using dynamic feeds (even if you’re not an ecommerce brand!): You don’t need to be an ecommerce business to utilise dynamic feeds.You might be selling cars, houses, or be a restaurant chain. Your stock, inventory, physical location, etc. can function as a “product” within a dynamic product feed so that campaign algorithms can scale or pause media spend in real time based on actual availability.
Using competitive price bidding: Campaigns shouldn’t be optimised in a vacuum. Live price comparison data can be fed into ad engines to automatically adjust bidding or pause spend when your product pricing becomes uncompetitive.
Feeding in macroeconomic signals: Campaigns can be fed external market variables that directly impact the demand for your product. Localised weather, economic conditions, consumer sentiment, regional demand fluctuations are amongst the many data sources that can shape campaign messaging, geographic targeting and budget deployment.
The data maturity spectrum
Typically, we see that brands sit in one of three tiers when it comes to data maturity. Those in tier 3 are the ones who will be seeing true net revenue growth. Those in tier one will likely be stuck with “fantastic” campaigns that don’t impact P&L.
Bridging the gap between surface campaign success and actual profitability requires moving up the data maturity spectrum:
| Data Maturity Tier | Input | Platform Behaviour | Business Outcome |
|---|---|---|---|
| 1. Low Maturity | Gross leads, total order counts, or raw conversion rates. | Chases volume without evaluating lead quality or transaction margin. | High budget waste; conversion metrics look strong on paper while net profit remains flat. |
| 2. Moderate Maturity | Value-Based Bidding (VBB) using static proxy values (e.g., arbitrary value rules for button clicks or calls). | Assigns relative weights to user actions, but relies on proxy estimates rather than real financial outcomes. | Improved value distribution, but remains disconnected from actual contribution margin. |
| 3. High Maturity | Real-time backend CRM data, margin inputs, and offline conversion tracking (OCT) via API. | Restricted signal sets force the algorithm to bid higher only on high-margin transactions or net-new customers. | Optimised bottom-line margins and sustainable long-term business growth. |
Moving up this spectrum changes the fundamental instructions you give to campaign algorithms. As shown here, when you sit at low data maturity, the engine will only see and optimise against low value conversion events. It will find the easiest route to success to hit that gross revenue target and will reharvest existing buyers or push zero margin products.
High maturity changes the game. Feeding real time CRM, margin, and macroeconomic data back into the system forces the algorithms to optimise for bottom line net profit, and that is precisely what will deliver true incremental revenue growth.
Brands need to take control of the machine. Tell it what actually drives the bottom line, and it will continue to chase it.
The bottom line: proper data integration into marketing delivers rapid, profitable, incremental growth
There is a reason that data is one of our core disciplines, and why our operational model, The Open Partners Operating System (OP:OS) starts with data. We won’t touch media or creative without first understanding what data is available, what it says, and how that data will be used to deliver against the clients’ ‘North Star’ metric (i.e. the key outcome that actually delivers growth).
In the algorithmic age of marketing, data isn’t an isolated technical capability, but a business critical, strategic necessity that, when applied right, unlocks real growth for brands.
