Measuring up

Client services, Operations and Marketing walk into a bar. Then Strategy. And Media. Creative arrives fashionably late. Maybe they are joined by a few others they texted along the way. Everything is great until it is time to settle the tab: “hey, has anyone seen Measurement?”

Let’s be honest. We’ve all been there. Measurement is tricky.

Technically, because we need to reconcile different activities happening in different channels at different levels across a decentralized stack. Operationally, because it is ill-suited to standard ways of working. In waterfall models, it is usually at the very end of a chain. As such, it is firmly on the tactical and implementation end, removed from the business and strategic decision-making. Organizationally, it seems to both touch everything and sit nowhere. As a result, the tools, expertise and governance are split between and within clients and agencies.

Yet, as we all know, none of this stops us from measuring what we do. We do it well, too. Over time, though, this state of affairs leads us to plan based on what and how we can measure. Ultimately, it sends us into a doom loop of pigeonholed strategic thinking and measurement strategy.

To escape the doom loop we must change the perspective on measurement from a cog in the machine to symbiotic scaffolding. In other words, to take measurement from a stand alone discipline to a part of an overarching data strategy stretching from client to agency and back. Through this lens, measurement both reflects and informs strategic choices.

From measurement doom to measurement boom

Measurement reflects strategy to the extent it helps us see whether we’re running things well, whether we’re running the right things, and whether they’re having the intended impact. It informs strategy when it contextualizes metrics and indicators, enabling us to quickly connect the dots, drill down on specifics or ask follow up questions of the data. At this point, what we have is a feedback-feedforward system that amounts to a boom loop.

Even though getting to this stage will require long-term commitments at some point, this virtuous cycle can be kickstarted by a shift of mindset and some tweaks in the ways of working. We can think about it as incremental steps, each with increasing commitments and rewards.

Step 1: Make every bit count Every bit counts for measurement.

Our choices are naturally encoded in parameters like audiences, channels, messages and orchestration rules. Each indicator of interest has a role in measuring the parameter. Whether it’s precision, reach, relevance, persuasion, cost-effectiveness, or any other, every one of them sheds light on a different aspect of the parameter. And for any given indicator, precision and accuracy are functions of one or more suitable metrics.

By explicitly connecting metrics to indicators and indicators to parameters, it is clear from the outset the answers we need to get, which indicator represents these answers and which raw metrics are needed to arrive at the indicator.

Step 2. Integrate measurement at different levels.

If we think about a project, its business, technical and operational domains are just different levels of the same thing. Therefore, we have to measure them all to have a complete picture. In our day-to-day, it means that business goals, campaign goals and operational targets should be explicitly aligned and telling different sides of the same story.

At the operational level we measure how efficient we are at creating, delivering and running our assets at intended channels and levels as per operational/SLA metrics (forecast x delivery, time to air, etc.). The operational/SLA metrics should go beyond media and include everything from admin to production. This is about watching out for leaks in the plumbing.

The technical level looks at marketing comms’ core activities. It measures how effectively we’re getting the message across. Examples of indicators include conversions, awareness, favorability, brand linkage, consideration and key message understanding, among (many) others. It measures the quality of the water flowing through our pipes.

The business level is the big picture. It should measure how effective comms is at driving the desired outcomes. It’s often represented by hard indicators such as transactions, hand raising, etc. In the end, it’s all about whether people will drink the water we are delivering.

Integration is understanding how each connects to and impacts the others and the whole. An integrated measurement should capture the interplay between the levels. With that in place, we can quickly know when (and where) problems arise, have the information to course correct and be precise in our interventions. Conversely, we can also quickly learn what is working and try and optimize. An integrated view lets us know at once – and at all times – if it is being well-run, if it is getting across and if it is moving the needle.

Step 3. Fold measurement into Data Strategy

Integrating measurement into data strategy means making it systemic. At this point, the alignment of goals (step 1) or domain level (step 2) to measures is less about project workflows and more about policy and governance. It’s how organizations think and execute measurement, and how it connects to the wider data ecosystem.

Here, marketing and comms data are seen as a valuable resource to be harnessed and as part of critical infrastructure that has to be managed and maintained. More often than not, the reality is that of multiple platforms (e.g. CRMs, ERPs, CDPs, DAMs, etc) from different vendors, each of which with their own (siloed) data models within their own (insular) environments. In addition to that, modern marketing comms involves multiple teams within the companies and their myriad external partners. Given this, how can it be ever folded into the data strategy?

What makes it feasible is measurement governance. Through it, for example, stack management, data roles, decision rights and accountability can be defined and assigned across the different internal and external teams, regardless of budget ownership. This is something all companies know how to and do across mission critical activities. It just doesn’t normally extend to marketing and comms measurement.

In this case, it means different things. First, putting in place data models to pull the data from third-party platforms and reconcile them into measurement frameworks that make sense for your business – i.e. your definition(s) of conversion(s) – and your data ecosystem. Then, the stack must not only store the data, but do so in a way that is both f it for marketing and comms’ specific purposes (e.g. graph databases for insight) and connects to other parts of the stack. With governance in place to guide all parties involved, the realities of marketing and comms activities can be met through documentation, procedures, APIs and other resources available to teams and partners.

The value of seeing measurement as data infrastructure

Most marketing and comm teams have campaign metrics dashboards. Fewer have a way to connect those to business impact. Almost none can tell you if the operational metrics that supported the campaign were even met. That is what measurement looks like as a stand alone discipline.

A measurement mindset focused on reflecting and informing goals and choices is the cornerstone of a sustained and sustainable virtuous cycle. It delivers value by reducing uncertainty, maximizing results, unveiling opportunities, mitigating pitfalls and enabling better decision-making. It makes the most of our resources by allowing for any point in the journey to become a metric, as long as we know what it means and the role it plays, both individually and within the wider frameworks.

The first step, making every bit count, can be easily deployed in campaigns just by getting the measurement people in the room early and keeping track of connections between goals, indicators and metrics.

Step 2, integrating measurement at different levels, requires more coordination as it demands input and alignment from the stakeholders involved in making a project happen. It also asks strategic-first people to embrace some technical details and technical-first people to get the strategic side. It also needs to account for different requirements and time scales. Operations might be closer to real-time tracking and have specific SLA caps to trigger warnings, for example, whereas technical measurement might follow journey and funnel performance at longer intervals (e.g. weekly) to capture trends with more certainty. Because of this, it benefits from greater control over project/campaign data (e.g. centralizing data from the many different platforms being used) and established analysis workflows (e.g. from cleaning to reconciliation to statistics).

Folding measurement into Data Strategy, step 3, is what takes things to the next level. This is where the measurement feedback-feedforward system truly becomes self-reinforcing. Through governance and the resources it makes possible, measurement’s full rewards can be unlocked without compromising timelines, introducing bottlenecks, or stifling stakeholders’ autonomy.

None of these steps will (nor should) happen overnight. But any of them can begin today. The key is having measurement at the table from round one. Otherwise the joke may end up being on us.

Paulo Coelho Mendes


Designs data products and ecosystems that unlock the power of ‘hard’ and ‘soft’ data, leveraging a mixed background in Sciences and Humanities.

Feel free to drop me a line at paulo@serendipitylab.co.uk to continue the conversation