Marketing teams have more ways to measure performance than ever before. They can track how customers discover a brand, interact with its content, click on an advert, visit a website, respond to an email, and eventually make a purchase. With multi-touch attribution, these interactions can be brought together to show how different marketing channels contribute to a conversion. In theory, this creates a more complete picture of the customer journey. In practice, it can introduce a different problem: more data points do not necessarily produce more accurate conclusions.
For many marketing leaders, attribution has become an exercise in assigning credit. A customer sees a social media advert, reads a blog post, clicks on a search result, opens an email, and eventually converts. Each interaction becomes part of the reporting process, and the marketing team attempts to determine how much credit each channel deserves. The resulting dashboard may appear sophisticated, but the numbers can still leave an important question unanswered: what actually caused the customer to convert?
This distinction matters because measuring the interactions that precede a conversion is not the same as measuring the incremental impact of marketing. A channel can appear frequently in customer journeys without being responsible for generating additional demand. Equally, a channel that rarely receives direct credit may play an important role in creating awareness, building trust, or influencing a customer's decision long before the final conversion takes place.
Multi-touch attribution is not inherently flawed. It can help marketers understand customer journeys, identify patterns, and examine how different touchpoints work together. The problem begins when attribution models are treated as definitive evidence of marketing effectiveness rather than as analytical tools with specific assumptions and limitations.
For CMOs and marketing leaders, the objective should not be to collect more attribution data or build increasingly complicated dashboards. It should be to understand how marketing influences customer behaviour, where investment creates additional value, and how confidently the organisation can connect its marketing activity to commercial outcomes.
The Promise of Multi-Touch Attribution
Traditional attribution models often assign conversion credit to a single interaction. First-touch attribution gives credit to the first recorded touchpoint, while last-touch attribution assigns it to the final interaction before conversion. These approaches are relatively straightforward, but they can overlook the influence of other activities that occur throughout the customer journey.
Multi-touch attribution attempts to address this limitation by distributing credit across multiple interactions. Depending on the model, a conversion might be credited to the first touchpoint, the last touchpoint, or several interactions in between. Some models assign equal weight to every recorded interaction, while others give greater importance to particular stages of the journey, such as the first interaction or the touchpoints closest to conversion.
The appeal is understandable. Modern customer journeys rarely follow a simple, linear path. A prospective customer might discover a brand through social media, return through an organic search, engage with a newsletter, compare alternatives, and convert days or weeks later. Looking only at the final interaction can overlook the earlier activities that helped bring the customer into consideration.
Multi-touch attribution gives marketing teams a framework for examining these journeys more comprehensively. It can reveal which channels frequently appear before conversions, how customers move between touchpoints, and where different marketing activities seem to contribute to the buying process.
However, the model's usefulness depends on the quality of the underlying data, the assumptions used to distribute credit, and the questions the organisation is trying to answer. A more detailed account of the customer journey does not automatically establish the causal impact of every interaction within it.
More Touchpoints Can Create More Confidence Than the Data Deserves
One of the biggest risks in marketing analytics is confusing measurement detail with measurement accuracy. When a dashboard records multiple interactions for every conversion, the resulting report can create the impression that the organisation understands precisely how its marketing investment produces results.
Yet every attribution model makes assumptions about how credit should be distributed. A linear model assumes that recorded touchpoints deserve equal credit. A time-decay model gives greater weight to interactions closer to conversion. A position-based model gives additional weight to selected stages of the journey. These approaches can produce different conclusions from the same customer data because they define contribution differently.
Consider a customer who discovers a brand through a paid social campaign, reads an article through organic search, receives an email, and finally converts after clicking a paid search advert. A linear model might distribute credit evenly across the four interactions. A time-decay model might assign more credit to the email and paid search interactions. A last-touch model would attribute the conversion entirely to paid search.
Each model produces a different account of the same journey. None can establish, from the sequence alone, what would have happened if one of those interactions had never occurred.
This is where marketing leaders need to exercise caution. Attribution reports can be useful for understanding recorded behaviour, but the allocation of credit is not necessarily a measurement of incremental impact. The model's output should therefore be interpreted in the context of its assumptions rather than accepted as an objective ranking of marketing effectiveness.
Attribution Measures the Journey, Not Necessarily What Caused the Conversion
A central limitation of multi-touch attribution is that it describes observed customer interactions more readily than it establishes causation. Knowing that a customer interacted with several marketing channels before converting does not prove that each interaction changed the likelihood of conversion.
Some customers may already have a strong intention to purchase before encountering the marketing activity that receives attribution credit. A customer searching for a specific product by name, for example, may be close to making a decision. A paid search advert might appear immediately before the purchase and receive substantial credit, even though the customer might have completed the transaction through another route.
Similarly, customers who engage with several marketing channels may already be more interested in the brand than those who do not. Their higher likelihood of converting could reflect their existing preferences, purchasing intent, or previous experiences rather than the cumulative effect of every recorded interaction.
This creates a potential problem when attribution results are used to guide budget allocation. If the organisation assumes that every channel receiving credit has generated additional demand, it may increase investment in activities that primarily capture existing demand rather than create new opportunities.
The distinction is commercially important. Capturing existing demand can be valuable, and conversion-focused channels have a legitimate role in marketing strategy. However, the business needs to understand whether an investment is creating new demand, improving conversion efficiency, retaining customers, or simply making it easier to complete purchases that were already likely to happen.
These are different outcomes, and they require different approaches to measurement.
The Attribution Model Can Change the Budget Recommendation
Marketing attribution is often treated as a technical exercise, but its consequences are commercial. The way a model assigns credit can influence which channels receive more investment, which campaigns are reduced, and how marketing performance is explained to the board.
Imagine a business running paid search, paid social, email marketing, and content marketing. Its attribution dashboard shows that paid search contributes the largest share of recorded conversions, while content marketing receives relatively little credit. Based on these results, the business decides to increase search spending and reduce investment in content.
The decision might be reasonable, but the attribution report alone cannot establish that it is the right one. Search may be capturing demand generated by other marketing activities, while content may be helping prospective customers discover the brand or evaluate its offering earlier in their journey. If the organisation reduces content investment without examining these relationships, it could weaken future demand while temporarily maintaining the appearance of strong conversion performance.
The reverse can also happen. A channel may receive substantial attribution credit because it appears across many recorded journeys, even when its incremental contribution is limited. Increasing investment based solely on that credit could produce diminishing returns.
This is why CMOs should be careful about allowing an attribution model to become the sole authority on budget allocation. Its findings should be considered alongside acquisition costs, customer quality, conversion behaviour, retention, revenue, margin, and evidence of incremental impact.
The objective is not to find the model that produces the most attractive channel rankings. It is to understand which investment decisions are most likely to improve the organisation's commercial performance.
Data Quality Can Limit the Value of Even the Most Sophisticated Model
Attribution models are only as reliable as the data they can access and connect. In many organisations, customer interactions are recorded across advertising platforms, analytics tools, customer relationship management systems, email platforms, and sales databases. These systems may use different identifiers, reporting windows, definitions, and methods of recording conversions.
As a result, the journey represented in an attribution dashboard may be incomplete. Some interactions may not be captured, while others may be duplicated or incorrectly connected to the same customer. Cross-device activity, consent restrictions, browser limitations, offline sales conversations, and gaps between marketing and sales systems can all make it difficult to reconstruct a complete customer journey.
The challenge becomes more complicated when different platforms report on the same conversion. Advertising platforms may each claim credit for a purchase because the customer interacted with several campaigns before converting. When those platform-level reports are combined without a consistent measurement framework, the organisation risks overstating the contribution of its marketing activity.
This does not mean the data is useless. It means the organisation needs to understand what each system measures, how the data is collected, and where gaps or inconsistencies may affect the conclusions.
A CMO should therefore treat measurement infrastructure as part of the attribution strategy rather than as a separate technical concern. Consistent conversion definitions, reliable tracking, clear reporting windows, agreed source-of-truth metrics, and stronger connections between marketing and sales data can improve the quality of the analysis.
Better data will not eliminate every attribution limitation, but poor data can make those limitations considerably more difficult to manage.
The Customer Journey Is Not Always a Linear Sequence
Another challenge with multi-touch attribution is that customer journeys are rarely as neat as reporting models suggest. A customer may see an advert without clicking it, discuss a product with a colleague, attend an event, read independent reviews, return through a direct visit, and then make a purchase after speaking to a salesperson.
Some of these interactions may be recorded, while others remain invisible to the organisation's measurement systems. Even when they are captured, their relative influence may be difficult to determine.
The problem is not simply that the business lacks enough data points. It is that the recorded interactions may not represent the full range of influences behind a purchasing decision.
This is particularly relevant for brands with long consideration periods, complex sales processes, or significant offline activity. A potential customer might encounter a company several times before actively engaging with its marketing. By the time the customer clicks an advert or fills in a form, important decisions about trust, relevance, and preference may already have taken place.
For these organisations, a model built primarily around trackable digital interactions can produce an incomplete view of marketing's contribution.
CMOs should therefore avoid assuming that every customer journey can be reconstructed from digital touchpoints alone. The reporting framework needs to reflect the realities of the buying process, including the influence of sales conversations, customer experience, brand familiarity, referrals, and other factors where relevant.
Replace the Search for One Perfect Model With a Broader Measurement Framework
The answer to attribution's limitations is not necessarily to abandon multi-touch attribution or replace it with another single model. A more useful approach is to combine different measurement methods, each designed to answer a specific business question.
Multi-touch attribution can help marketing teams understand observed customer journeys and examine how channels interact. It is useful for identifying patterns, investigating conversion paths, and diagnosing changes in performance. However, it should not be expected to answer every question about marketing effectiveness.
Other methods can provide complementary evidence.
Marketing mix modelling can help estimate the relationship between marketing investment and business outcomes across channels and over time, while accounting for other relevant factors where the data and model permit. Incrementality testing, including controlled experiments and suitable holdout tests, can help estimate whether a marketing activity generated outcomes that would not otherwise have occurred. Cohort analysis can help marketers understand how customers acquired through different channels behave after conversion, including differences in retention, spending, and long-term value.
These methods have different data requirements, costs, strengths, and limitations. They are not interchangeable, and none should be treated as a universal solution. Their value comes from answering different questions and providing evidence that can be considered together.
For example, attribution data might suggest that paid search contributes significantly to conversions. An incrementality test could then investigate whether increasing or withholding a particular part of that investment changes total conversions. Customer cohort analysis could reveal whether customers acquired through the channel remain active and generate sufficient long-term value.
Together, these perspectives can provide a stronger basis for investment decisions than attribution credit alone.
Incrementality Helps Answer the Question Attribution Cannot
If attribution examines how credit is distributed across recorded interactions, incrementality focuses on a different question: what additional outcome did marketing generate?
This distinction can change how marketing performance is evaluated. Rather than assuming that every attributed conversion was caused by the channel receiving credit, the organisation investigates what would have happened without the marketing activity in question.
A controlled experiment might compare an exposed group with a suitable control group to estimate whether a campaign produced additional conversions. Depending on the channel and business context, marketers might also use geographic experiments, audience holdouts, or other testing approaches designed to establish a credible comparison.
The results can help reveal whether a campaign is generating additional demand, improving conversion, or primarily reaching customers who would have converted anyway.
However, incrementality testing also requires careful design. The comparison groups must be appropriate, the test must account for relevant external influences, and the measurement period must be suitable for the outcome being assessed. Poorly designed experiments can produce misleading conclusions just as easily as poorly interpreted attribution reports.
Not every channel can be tested in the same way, and some campaigns may require alternative approaches because of their scale, timing, or influence across the wider market. The objective is to select the strongest practical method for the question being asked.
For CMOs, incrementality is especially useful when making decisions about budget expansion, channel saturation, promotional activity, and investments that appear successful under attribution but may not be generating sufficient additional value.
Revenue Is Important, but Customer Quality Matters Too
Even when attribution appears to connect marketing activity to conversions and revenue, the analysis remains incomplete if it ignores what happens after acquisition.
Two campaigns may generate the same number of customers and similar initial revenue while producing very different commercial outcomes. Customers acquired through one campaign may retain longer, purchase more frequently, generate stronger margins, or require less support. Customers from another campaign may convert quickly but disengage soon afterwards.
If the marketing team evaluates both campaigns only through attributed conversions, it may overlook these differences and allocate budget towards acquisition that looks efficient initially but produces weaker long-term value.
This is why attribution should be connected to customer and commercial metrics. Customer acquisition cost, customer lifetime value, retention, payback period, gross margin, repeat purchase behaviour, and revenue quality can help the organisation assess whether marketing is attracting customers who contribute meaningfully to business performance.
The appropriate metrics will depend on the organisation's business model. A subscription business may prioritise retention and recurring revenue, while an e-commerce business may examine repeat purchase rates and contribution margin. A business with a longer B2B sales cycle may focus on qualified pipeline, sales conversion, deal value, and the time required to recover acquisition investment.
The underlying principle is consistent: a channel's value cannot be fully understood by examining how much conversion credit it receives. Marketing leaders need to understand the quality, cost, and durability of the outcomes associated with that activity.
The Best Attribution Reports Explain Uncertainty, Not Just Results
One of the most important responsibilities of a CMO is to communicate what the data can and cannot establish. Executive reporting becomes more useful when it distinguishes between observed performance, model-based estimates, and conclusions supported by stronger causal evidence.
For example, a report might show that paid social received 25% of attributed conversion credit during the reporting period. That is a description of the model's output. It does not necessarily mean paid social generated 25% of the conversions or that removing the channel would cause conversions to fall by the same proportion.
A more useful report would explain which attribution model was used, what the result suggests about the customer journey, how the finding compares with previous periods, and what additional evidence is available to support an investment decision.
Where an incrementality test has been conducted, the report can distinguish its findings from the attribution results. Where no such test exists, the CMO can identify the uncertainty and explain how the team intends to investigate it.
This approach does not weaken the marketing function's credibility. It demonstrates that the team understands the difference between measurement and interpretation.
A strong executive reporting structure should therefore follow a clear logic: what the data shows, what it suggests, what remains uncertain, why the finding matters commercially, and what action the organisation should consider next.
This gives the board a more realistic understanding of marketing performance and helps prevent estimates from being presented as established facts.
Build an Attribution Framework Around Business Decisions
CMOs looking to improve attribution should begin with the decisions the organisation needs to make rather than the capabilities of the analytics platform already in use.
A practical framework is to:
- Define the business question: Establish whether the organisation needs to understand customer journeys, compare channel performance, estimate incremental impact, or evaluate long-term customer value.
- Choose the appropriate measurement method: Use multi-touch attribution for journey analysis, experiments for suitable incrementality questions, marketing mix modelling for broader investment analysis, and cohort analysis for post-acquisition behaviour.
- Establish consistent data definitions: Agree on how conversions, customers, revenue, acquisition costs, and other key outcomes are recorded across systems.
- Understand the model's assumptions: Document attribution windows, credit-allocation rules, data limitations, and the circumstances in which the results may be unreliable.
- Connect attribution to commercial outcomes: Examine customer quality, profitability, retention, and payback alongside recorded conversions and channel credit.
- Validate important investment decisions: Where practical, use experiments or other independent evidence to test whether attributed performance reflects additional business value.
- Report uncertainty clearly: Distinguish observed data from estimates and explain where further evidence is needed before making significant budget changes.
- Turn findings into action: Use the combined evidence to guide budget allocation, campaign optimisation, testing priorities, and future investment decisions.
This framework helps marketing teams move beyond the assumption that a more sophisticated attribution model will automatically produce better decisions. It places the emphasis on choosing the right measurement approach for the question at hand and ensuring that the findings remain connected to commercial priorities.
Conclusion
Multi-touch attribution can provide valuable insight into how customers interact with marketing channels before converting. It helps marketers examine customer journeys, identify patterns, and understand how different activities appear to work together. However, the ability to track more interactions does not automatically establish which activities caused a conversion or generated additional business value.
The risk emerges when the credit assigned by an attribution model is treated as a definitive measure of marketing effectiveness. This can lead organisations to overinvest in channels that capture existing demand, undervalue activities that influence customers earlier in their journey, and make budget decisions based on assumptions the underlying data cannot fully support.
For CMOs, the priority should be to build a measurement framework that combines customer journey analysis with commercial performance, incrementality testing, and other appropriate methods. The objective is not to find a single model that answers every question. It is to develop a clearer understanding of what marketing is contributing, where investment is producing additional value, and how confidently those conclusions can inform business decisions.
The strongest marketing leaders understand that measurement is only useful when it improves decision-making. They know when attribution provides a helpful signal, when further investigation is required, and when a different method is needed to establish the true impact of an investment.
More data points can make the customer journey easier to observe. Better questions, stronger measurement, and disciplined interpretation are what make marketing performance easier to understand.
If your marketing dashboard can explain which channels received credit for a conversion but cannot tell you whether those channels generated additional value, it may be time to rethink what your attribution model is actually designed to measure.
Join the Growth Authority Community
If you're a CMO or marketing leader, we've created a community called Growth Authority to help you stay ahead of the curve. Inside, you'll gain access to proven frameworks, dashboards, and tested playbooks to help you lead confidently in this new era of marketing. This private community is built exclusively for senior marketers who want early access to cutting-edge strategies before the rest of the market.
Be among the first to access these resources: Join the Waitlist.
Make the opening more compellingReduce repeated attribution explanations

