Marketing teams are under increasing pressure to demonstrate the value of every pound spent. As budgets face greater scrutiny, businesses are looking closely at their technology investments, questioning which platforms are essential and where costs can be cut.
There is good reason for this reassessment. Gartner’s Marketing Technology Survey found that organisations use just 49% of their martech capabilities. For businesses paying for technology they aren’t using, streamlining the stack can seem an obvious way to improve efficiency.
But cutting measurement capability is not the answer, and can prove more costly in the long run. Reducing the number of tools may look like progress, but it doesn’t solve the underlying challenge. A business can remove several platforms and still be left with the same fundamental problem: disconnected data that makes it difficult to understand performance, compare results and demonstrate return on investment.
Rather than simply cutting tools, businesses should focus on streamlining their processes. The tech stack shouldn’t be a ladder to climb: when adoption is straightforward, productivity improves. That doesn’t mean having more tools or fewer, but making sure each one serves the right purpose and helps give AI the context it needs to read and analyse data the way the business needs it to.
Fewer platforms, same blind spots
Marketing measurement has become increasingly complex as businesses invest across more channels, platforms and customer touchpoints. Each system generates valuable information, but that information is often collected and interpreted differently.
Take a marketing team measuring customer acquisition across display, paid search and social media. Each platform may report conversions according to its own attribution methodology, while the business’s customer relationship management system records customer outcomes using another set of rules.
Individually, these reports can provide useful insights. But bringing them together to understand which investments deliver the greatest value is considerably harder.
Removing one of those platforms doesn’t resolve the differences between the remaining datasets, nor does it establish which metrics the business should prioritise when evaluating performance.
This is why technology consolidation and measurement improvement need to be treated as related but distinct objectives. Businesses should assess their existing capabilities, identify genuine duplication and understand how their data connects before deciding what to remove.
What counts as a conversion?
One of the biggest barriers to effective measurement is the absence of shared definitions across marketing data. Metrics such as conversions, customer acquisition cost and return on investment may appear consistent on the surface, but the calculations behind them can vary significantly between platforms and departments.
One team might define a conversion as a completed purchase, while another includes newsletter sign-ups or other customer actions. Both definitions may suit their respective objectives, but combining them without recognising the difference creates a misleading picture of performance.
The solution starts with a common understanding of what important metrics really mean, where they come from and which outcomes they represent.
That requires collaboration between data, finance and marketing teams. Marketing understands campaign objectives, finance brings the commercial perspective and data teams can make sure definitions are applied consistently across systems.
With these foundations in place, businesses can compare performance more reliably and make investment decisions based on a shared understanding of the results.
Right numbers, wrong conclusion
These foundations matter even more as AI takes on a larger role in marketing measurement.
AI can help teams analyse large volumes of information and investigate potential opportunities much faster than manual reporting. However, the reliability of those insights depends on what the technology understands about the data it is analysing. Clean, well-structured data is essential, but it isn’t sufficient on its own.
An AI system might identify a sudden decline in campaign performance and recommend reducing investment. Yet that decline could reflect a deliberate change of approach, such as a temporary budget adjustment or a shift towards reaching new audiences rather than driving immediate conversions.
Without that business context, the system could interpret the numbers correctly while still drawing the wrong conclusion.
AI therefore needs a governed understanding of how metrics relate to one another, alongside the commercial context that explains why performance is changing. At Adverity, we call this the marketing knowledge layer: the agreed metric definitions and the business context behind them. With it, AI can read data the same way the business does, removing the ambiguity.
The distinction matters because a recommendation can appear convincing even when the assumptions behind it are wrong. As businesses rely more on AI to support budget allocation, those errors could have direct financial consequences.
From reconciling figures to making decisions
The real purpose of measurement is to understand what’s working and decide what to do next.
Too often, valuable time is spent gathering information from different platforms, reconciling conflicting figures and preparing reports before meaningful analysis can begin.
Giving marketers easier access to consistent, reliable insights can change that. Teams can investigate emerging performance issues, understand what is behind them and respond while there is still time to improve outcomes.
AI can accelerate this by making complex analysis more accessible, but it should support human judgement rather than replace it. Marketers still need to decide which outcomes matter, weigh recommendations against commercial priorities and judge where investment will create the greatest value.
For businesses reviewing their technology investments, the starting point should be measurement capability rather than platform count. Cutting unnecessary technology is sensible when budgets are constrained, but it’s only an improvement if it also helps the business operate more effectively.
Before cutting a tech tool, ask one question: will we still be able to see what’s working?
