Marketing Attribution Models: The Good, The Bad & The Best For Your Team

Here’s the hard truth about marketing attribution: it doesn’t show you where your budget is most effective. In fact, most attribution models are actively damaging your team’s ability to measure advertising impact on customer acquisition - and therefore your ability to deliver and demonstrate ROI from £millions of media spend. 

CMOs and Marketing Leaders are well-versed in media strategy - but is your current model really maximising revenue growth? Effectiveness Leads are grappling with so many channels, datasets and stakeholders that the customer journey has taken a backseat to short-term vanity metrics. The best marketing strategies are only as successful as the teams behind them. Many teams are working under a false sense of security with multi-touch attribution models that sound perfect, but in reality, are not underpinned by the right data - and will therefore not deliver the right results. 

World-leading CMOs and industry experts share the hard lessons they learned from implementing media measurement, and how your team can choose - and succeed with - the right marketing attribution model to drive revenue for your business.

The Myths, Truths and CMO-Proven Strategies Behind Every Single and Multi-Touch Attribution Model

Model 1: First-Touch Attribution - The Perils of the Lead Generation Obsession

The Good - The longest and most complex stage of the buyer’s journey, awareness is accepted to be the least understood by marketing teams. To drive enough volume through the funnel from the beginning, the awareness stage often sees the highest output of content. However, CMOs know much awareness content is inadequate at converting audiences because:

• Few marketers are devoting sufficient time to the promotion/production 80/20 ratio of the Pareto Principle

• Copy writing and design isn’t fully harnessed for sufficient content atomisation 

• PPC, organic social and content teams are not regularly reviewing and analysing performance, and are not collaborating enough to see the full picture of each element of the marketing mix and combined as a whole

• Content isn’t good enough to start with, either or both in terms of general quality and close alignment to customer pain points.

Some marketers argue that first-touch aids more effective awareness tactics by highlighting what works earlier on, from channels and communication styles to actual creative and copy itself. Marketers can use awareness stage learnings not only to improve individual campaigns in-flight for optimum efficiencies and effectiveness, but also to build better customer understanding. Awareness stage insights on customer pain points, behaviours and intent makes for better informed marketers generally, who can apply insights to other funnel stages, other campaigns and the overall marketing strategy. 

The Bad - Fundamentally, first-touch lies. Often. In theory, first-touch can highlight which channels are best at converting audiences. In practice, the model doesn’t take into account the many activities and instances of intent that must have come before the touchpoint. The model isn’t really ‘first-touch’, but the first touch you can see. This model assumes that the most obviously trackable touchpoint has the biggest impact on the conversion, completely disregarding consumer intent and behaviours at every other invisible touchpoint in the dark funnel, before the consumer decides to visibly interact with a brand. 

The age-old battle between sales and marketing has much to answer for first-touch attribution mistakes. The very concept of ‘leads’ distracts focus and diverts resources away from the best deal opportunities. Sales complain about lead volume and quality, and marketing complain that sales are not following up on leads. If lead quality does need improvement as claimed, sales teams are often failing to provide feedback and insights to enable this improvement from marketing.

If the ‘first’ attributable touchpoint is actually the first time a lead has been touched, this model is sending individual consumers to sales who aren’t anywhere near ready to convert. If they are ready to convert, this touchpoint is not really ‘first’ because the consumer in question will have been captured and nurtured by other channels, content and comms. First-touch attribution both jars with the concept of lead generation and perpetuates the short-term obsession to grab hold of any consumer demonstrating any intent whatsoever and force them down a faster, narrower buyer’s journey.

First-touch attribution is encouraging marketers to prioritise the wrong content in the mistaken overestimation of its conversion value, focusing the whole funnel down a path that is less relevant and helpful for customers. First-click focus delivers a poorer experience and, ironically, actually reduces the genuine conversion rate.

The Best: How to Implement the First-Touch Model for Better Insights - First-touch attribution will never present a full 360-degree customer view, but it can be a helpful tool when used as a smaller part of a well-planned and executed strategy to measure advertising ROI. Marketing teams must stop focusing on internal stakeholders and their demand for lead volume or (what they perceive to be as) return on investment: take the focus back to the customer. Invest in a better customer experience throughout the funnel, starting with better awareness stage content.

Najoh Tita-Reid, Chief Brand and Experience Officer at Mars Petcare, and one of Marketing Week’s Top 100 Marketing Leaders of 2023, reveals on the issue: ‘I am concerned that about two thirds of Gen Z just believe that advertisers are selling them more stuff that they don't need. We're losing trust, there's a belief that we're no longer really focused on driving real consumer value. Our industry is more at risk than it's ever been before. If we don't become value-additive, consumers have more power than ever before to cancel us or not share the data we need to stay engaged. We have to earn the right to be in their life, to get their attention, more than we've ever done.’

The pressure to deliver leads and conversions means marketers are rushing to sell to consumers too early, and with content that doesn’t help them solve their problems or add value to their work/lives. Invest in sufficient time and resources to promote high quality content, and analyse engagement to see which is most effective, to continuously improve the impact of content on conversion.

Mars Petcare refocused their content strategy to complement a good existing understanding of what shoppers liked in-store with improved understanding of online content effectiveness, creating opportunities for an even better customer experience across channels. Tita-Reid and her team prioritised customer understanding and content quality, with many successful tactics including:

• Partnering with Ipsos to deliver the world’s largest study of the relevant audience category (Pet Parents) and continuously investing in Human-Animal Interaction research with the Waltham Petcare Science Institute. These scientific endeavours headed by expert partners deliver unique, never-before-seen first-party research that presents the best possible understanding of customer lives, pain points and goals. Mars Petcare’s marketing team harness this knowledge to create and communicate the most powerful, engaging and original awareness content.  

• Investing in AI to test and learn, relying on human skill and customer knowledge to create the content in question, with the help of emerging technologies to experiment faster. Content teams use suggestions generated by Artificial Intelligence to make small adjustments to multiple content pieces every month. This aids understanding of audience response and the impact of each content piece in the buying journey. 

• Balancing content production volume and quality with creativity and effectiveness of promotion. Mars Petcare showcases customers as the heroes in their own stories and quickly capitalises on relevant trends of interest to their ICP - with consumers always at the heart of content narratives - across social media channels.

• Working closely with the CFO and finance department to agree and best allocate sufficient marketing budget. Tita-Reid has appointed a new digital agency and joined forces with other partners to invest in original content creation, distribution and promotion to maximise audience awareness and engagement.

Model 2: Last-Touch Attribution - Wasted Ad Spend is the Least of Your Worries

The Good: The clearly defined nature of last-touch attribution as an event immediately before a sale is easier to measure and implement. Last-touch means marketing teams don’t need a high volume of data to map the customer journey and brand interactions before making a purchase, a major time saver in a world where the average organisation uses 120 different martech tools.

Last-click analysis is often the default offered by advertising platforms, making it easier for PPC and SEO teams to analyse data in the same way across the marketing mix, and reducing margin of error in cross-channel reporting. A naturally shorter time gap between the most recent consumer touchpoint and the actual purchase supports advertising managers with easier campaign management and measurement of short-term sales cycle impact. In principle this model proves efficient for marketing teams who can best harness it to move consumers from consideration to decision stage. Teams can discover which content, comms and channels are most effective at converting customers. 

The Bad: Last-touch involves many of the same pitfalls as its earlier-funnel cousin. Isolating single events ignores the full picture of the journey taken to get to lead and conversion state, limiting and potentially even harming the marketing team’s understanding of customer intent and behaviours. A short-term focus again detracts and distracts from marketing’s number one purpose: driving a brand that powers long-term business growth. 

Whilst the brains of marketing leaders have been all but permanently stamped with Binet and Field’s 60/40 split of brand and activation as best practice, most of their teams are not following through. Advertising leads are choosing the path of least resistance instead of the best return: last-click narrows focus to easily-available metrics and immediate outcomes, regardless of their actual value. Short-term activation is currently receiving an eye-watering 80% of the average advertising budget, completely ignoring the mammoth 95% of customers not currently in-market.

Last-touch encourages lazy grabbing of the bottom of the funnel, which will ultimately fail to deliver anything beyond short-term sales uplift. PPC is often aimed at audiences who have already seen content and comms across other channels, who are already being nurtured towards purchase, and who may have gone on to purchase regardless. Pressure to deliver efficiencies and report ROI quickly means marketers are wasting their ad spend on the 5% of audiences most likely to buy. Ultimately, marketers aren’t marketing: they’re making easy sales, and delivering little value for their organisations.

Last-touch is also costing you customers - a lot of them. 65% of marketing departments rely on this attribution model despite Accenture’s alarming findings: last-touch misses out on over a third (35%) of all sales. Inaccurate and incomplete data informs poor selection of channels, creative and messaging, meaning that campaigns are both wasting money and resources targeting the wrong audiences, and ignoring the right audiences in the places where they do already exist. 

The ‘click’ bias of the model also unfairly favours some channels over others in the journey to purchase.  Almost 3 in 4 social media users discover new brands through one marketing channel, such as TikTok, but actually click to purchase through another channel like search, completely bypassing a narrow attribution model. Individual customer actions are missed across an increasingly complex buying journey. 

As some channels such as TV are inherently unclickable, their impact is downplayed, and ever-growing multi-screen and multi-channel usage are ignored. Recent research revealed that consumers who clicked on Google and Facebook adverts were exposed to 39% more TV ads in the week before clicking, meaning that the clicks to purchase were driven by TV advertising rather than the more easily measurable online PPC. Reliance on last-click behaviours is applying an old-fashioned perspective to an increasingly empowered and informed consumer audience. With a better buying experience, happy and engaged consumers can enhance marketers’ understanding of their needs and as a result help drive customer acquisition and loyalty. 

The Best: Don’t. Or Do, But Focus On Brand & Real Demand Generation - Panic over short-term ‘return’ is driving an obsession with granular attribution, making last-touch attribution the answer to an irrelevant question. Last-click metrics are only helpful when used as part of a much wider strategy that prioritises long-term brand growth with activation as a supporting tactic.

Your 95% of ‘out-market’ buyers will not click on a purchase CTA even if that option is available to them on the relevant channel. The vast majority of potential customers are better served by genuinely interesting and helpful content, alongside compelling brand advertising that builds personal connection over time. Marketing teams know all this (or they should) but need to refocus on distributing efforts where they add the most value across the customer journey.

Ben Davis, Insights Editor at Marketing Week, advises, ‘Think of brand terms in paid search – when a shopper has typed your brand name into a search engine, do they really need to be served a PPC ad in order to find your site when the organic listing is right there? Or a shopper who has already browsed a product, then gets retargeted around the web with display ads – if they then come back and buy, how much is really to do with those ad impressions?’ Recent research into UK supermarket advertising challenges the popular tactic of allocating significant spend to own branded keywords on Google search: effectiveness is questionable, as many users who click a paid link would have otherwise clicked the organic link without any cost to the brand.

Brand is and always will be the main driver of long-term business growth. Advertising teams must refrain from wasting ad spend on bottom-of-the-funnel guaranteed conversions, and instead get comfortable with slower but much more valuable return. 

Richard Bowden, No.7 Marketing Director and B2B CMO brand growth veteran shares, ‘Short-cutting the marketing process means that you could run the risk of missing something that's fundamental. This is where the process comes in: you can follow your workings back and reverse-engineer things. When I've worked on brands that need to be reignited - brands that need help to return to their old glory days - they’re often the ones that hadn't had enough work put into their mass marketing communications - or ‘the long of it’.

‘At some point in your marketing mission you need to get those sales conversions. You need patience because it can take time - but it will come, and you’ll start to drive that return. Because these brands have let the brand marketing slide, and focused more on sales activations, it takes around three times the investment to get the brand back to where it was. If the brand had maintained a careful balance of the long and short all the way through, it would be in a much healthier position.’

Model 3: Multi-Touch Attribution - The Myth of Defining Advertising ROI

The Good: Tracking the entire customer journey across multiple channels, multi-touch attribution facilitates a much better understanding of the customer as a whole and their engagement with each channel, content and comms piece in their journey. This model helps marketers operate with an agile approach, seeing which interactions are really driving customer acquisition forward across advertising, content, landing pages, apps and conversations with sales, and working to optimise accordingly.

Marketers can allocate their time and resources in the places that are generating most revenue, and learn from these to adapt and improve other channels and content. MTA not only optimises marketing resources but also those of sales, helping both teams to align more closely and zero in on the most influential interactions. More accurate reporting on the most important interactions also aids better sales reporting, providing leadership teams with a more comprehensive overview of ROI and where spend is best allocated to drive customer acquisition.

The Bad: Determining the influence of each channel is still not an exact science. The model is only as effective as the way it is used, namely how each touchpoint and its contribution in the customer acquisition process are defined, how sources are measured and analysed. Since the phasing-out of third-party cookies at the hands of Google and Apple, much of the data driving the multi-touch model will no longer exist. Marketing teams will only be able to see the last click in the journey of impacted audiences. The influence of earlier digital touch points, and therefore the credit they can be given for customer acquisition, will be at best skewed and at worst disregarded altogether.

Working with third-party advertising partners in a cookieless world also presents problems for users of the MTA model. Advertising teams are in danger of wasting spend on retail media networks who are serving ads to known audiences on low quality websites, whilst claiming that their source is effective at converting, and appearing to provide evidence that backs this up. Third party ads can appear a low-cost and effective alternative, but their view-based ROAS metrics don’t provide the full picture of the path taken to conversion, leading marketers to overvalue and select the wrong channels for investment. Wasted ad spend pales in comparison to the brand damage caused by poor quality ads, poor customer experience and missed opportunity to engage with audiences altogether.

MTA often undervalues channels and activities earlier in the funnel that may appear to deliver lower ROI in the short-term, but are much more powerful in driving long-term sales growth. Frontier report on the dangers of misleading attribution metrics, highlighting a recent case study where ROAS declined as a CPG brand optimised towards incremental sales on Amazon: ‘Ad spend increases and topline sales grow at an even faster rate, implying an increase in ad-spend efficiency (every dollar invested is generating more topline growth than it has previously). ROAS paints a very different picture, declining 17% during the same period.’

The Best: Partner Experts for a Long-Term Strategic Approach to Multi-Touch Attribution 

Partnerships with data and analytics experts are a route to a better picture of the customer acquisition journey and help drive long-term sales, as recently demonstrated by TikTok. The social media behemoth is aiming to improve customer understanding of the relationship between TikTok campaign performance and the bottom line. TikTok’s new partnership with information and insights specialist TransUnion will harness MTA within campaigns through privacy-first data collection processes that provide omnichannel insight. TikTok’s new solution will combine point of sale input with other information from inside the existing CRM, for a fuller view of how each campaign is performing and the impact of each on conversion.

Globally renowned CPG leader The Hershey Company is harnessing insights from partnerships with other market leaders like Disney, and collaborating with a long-term focus to create effective multi-touch attribution models. Vinny Rinaldi, Vice President of Media & Marketing Technology at The Hershey Company, reveals, ‘It took 18 months to get this through the door. Cheap scale earns higher ROI, but the other side of the equation is effectiveness: revenue due to that channel. We can lose sight of that other side because we look at just CPMs we lowered versus the more expensive media actually driving a greater effectiveness, which is more revenue to our top line. 

‘Reverse the equation: instead of focusing on a single number of ROI, change the narrative internally, talk to your leadership and walk them through this. For us it didn’t make sense to look at the cost of what a Halloween integration will look like because of the impact and the synergy it creates for our business. When you look at the power of Disney impressions and integrations, you get a two-times return on investment, as well as a three-times growth in effectiveness. This is top-line revenue during the most pivotal part of our year — which last year gave us our first billion dollar season in Halloween. When you start to really surround the moment, you start correlating and understanding what is really playing a role in driving revenue for your company.’ 

Media measurement partners don’t just manage attribution models, but create winning advertising effectiveness strategies - empower your team with the knowledge to drive revenue growth.

Types of Multi-Touch - More Complex Attribution Methods 

Multi-Touch Model 1: U-Shaped Attribution

The Good: As conversion value is split more equally across the whole customer journey, U-shaped attribution provides a much better understanding both of the consumer and their conversion path than first-touch or last-touch alone. The combined approach recognises the importance of the first trackable sign of intent and the specific action aligned with conversion, enabling advertisers to step back and take a more strategic view than with single-touch models.

This model gets around the intensive set-up and training required for more complex tracking and reporting, with even the most junior marketers able to work to the ready-made formula, U-shaped aids efficiencies and reduces user error.

The Bad: Assigning only 20% of conversion value to every touchpoint except the first and the last necessarily overvalues some touchpoints and undervalues others. U-shaped users encounter the same problems as other restrictive models, in that the majority of consumer interactions are disregarded. The content, comms and channels that contribute to conversion throughout the journey are not effectively reviewed and analysed.

U-shaped can lull marketers into a false sense of understanding of their market, and a false belief that they are operating strategically when their focus will not grow beyond tactics. This model appears to deliver a full picture view but can in practice cause misdirection of budget and application in greater volumes - and with more damaging results - than other models. 

Best Implementation: For very short sales cycles involving only a handful of interactions, or start-ups and small businesses looking to gain useful understanding of the funnel, U-shaped can provide valuable insights into what are most likely important conversion influencers. For every other circumstance, this model can be a powerful addition to an existing account-based marketing strategy.

Implement account-based attribution reporting that looks at the entire journey per account, instead of just individual leads, by collecting data from multiple sources across the entire funnel. Marketers can better understand the impact of both marketing and sales activities on each decision-maker in each account through strategic martech usage and tracking only the most relevant data across CRM, automation, analytics tools and channels. 

Multi-Touch Model 2: Linear Attribution

The Good: A simpler way to get started with MMA, assigning equal credit to all touch points is easier for marketing teams to manage and provides a much wider view of the path to conversion than just first-touch and last-touch measurement. Marketers can develop better understanding of how their channels are working together to move consumers down the funnel, and continue to improve their understanding because all touch points are fairly considered as potentially important. 

The Bad: Not all channels, content or ads are created equal, meaning the linear model naturally starts with data that isn’t entirely accurate, and continues to undervalue some actions and overvalue others. This method fails to account for the influence of specific activities, meaning marketers base optimisation on incorrect assumptions and waste ad budget. Insights are too generic, proving even less relevant for more diverse customer groups, and too simple for complex sales cycles. 

Best Implementation: Linear measurement is far more accurate when underpinned by optimal customer understanding. Marketing teams should revisit their customer persona, conducting further research and testing to map the buying journey in greater detail. Collate and analyse your data as a whole: run different calculations between first and last-click and other touch points to test alignment with your persona research.

A more accurate starting point strategically will reduce the impact of less accurate reporting of channel influence, and can help marketers experiment with more awareness to anticipate and work with any skewed results.

Multi-Touch Model 3: Time Decay Attribution 

The Good: Allocating more value to touchpoints closer to conversion provides a more comprehensive view of engaged audiences, and in theory of intent-driven interactions created by your more effective content and channels. Time decay is more flexible in nature and can further enhance understanding through easy integration with other channels for a fuller picture of the customer journey. 

Unlike other models, time decay benefits longer sales cycles, with sales teams able to better connect their efforts to those of marketing. Later-stage insights drive better understanding of customer decision-making, further positioning marketing as an integral partner to the sales team. 

The Bad: Awareness stage understanding, already perilously low amongst many marketing teams, is given no thought and will see zero improvement. The majority of potential customers are lost at the awareness stage, meaning an absolute prioritisation of later-stage interactions will directly impact sales and the bottom line. The accuracy of the model as a whole is debatable given the setup, variety of tools, time-stamping and timeline building required, and the continual need to review and refresh. 

Best Implementation: Users of this model must regularly and sufficiently analyse the whole conversion path to accurately measure the influence of both recent and earlier interactions.The flexibility of time decay works well when used alongside other more rigid and focused attribution models like linear to gain more detailed and accurate insights. Time decay also complements effective testing and learning strategies: teams that experiment with new tactics and channels, such as influencer and conversional marketing, can build better understanding to improve and personalise a customer experience that converts and delights. 

Multi-Touch Model 4: W-Shaped/Full-Path Attribution

The Good: W-shaped and full-path are so similar, with many of the same benefits, that both can be assessed together. On the surface, full-path is the attribution dream. Measuring the first and last touches, the lead conversion touch and the touch believed to create the opportunity, as well as other touch points along the journey, provides the widest customer journey view of all single and multi-touch revenue attribution methods. Full-path attribution recognises that post-lead interactions can be just as intent-driven as earlier engagement.

Both expansive models enable marketers to gain understanding of consumers as individuals on an individual buying journey. B2B teams especially gain new insights into engagement across smaller steps of a more intricate buying process, particularly a more varied consideration stage with much larger groups of stakeholders and purchase influencers.

The Bad: The level of detail required not only makes the model more difficult to use, but also widens opportunity for error and misguided interpretation. Defining specific touch points and meaningful engagement along the buyer’s journey is always a best-guess scenario without the right combination of datasets and insights.

Without the full picture of the customer journey, W-shaped and full-path still don’t take into account the many hidden digital, offline and in-store touchpoints that form crucial parts of the awareness, consideration and decision stages. Marketers are operating in the dark as long as they rely only on one singular model as the answer to every consumer engagement question.

Implement Properly: Consistent measurement and smarter use of marketing spend will provide a good starting point for Effectiveness Leads and advertising teams. Create a clear taxonomy with an overview of every pound, and minimise content production costs by experimenting with small changes to reusable assets.

Trend monitoring is vital to a test-and-learn approach that delivers meaningful insights. Advertisers should remain aware of the dangers of relying on emerging technologies, but with controls in place can harness machine learning and AI to better understand existing attribution methods.

Integrating a wide variety of data sources is vital to building the full, accurate and genuinely helpful picture of the customer journey. Support W-shaped or full-path models with regular, detailed and relevant first-party research into your customer audience. Speak with customers directly through focus groups, surveys and research reports, work closely with sales teams to create feedback loops, and continually seek out and utilise opportunities for offline and online customer interactions. 

At its heart, marketing is and always will be customer-focused - accurate media measurement must follow suit.

What Multi-Touch Attribution Can’t Do - Single-Source is the Advertising Impact Solution

The concept of MTA makes perfect sense: tracking every touchpoint on the customer journey to understand and improve the path to purchase. But there’s a problem: no source can provide the data marketing teams need to properly implement MTA.

Single-source data does everything that MTA promises but can’t deliver. A comprehensive collection of information, single-source data is gathered from a single person, revealing their entire journey to purchase across channels. 

Viewers Logic harnesses single-source to measure the person, not the media. We track TV and online media exposure, combined with online and offline purchasing behaviour, over time on the same individual, to form the full detailed picture of every customer.

The Most Effective Way to Measure Ad Spend Effectiveness - The Best Alternative to Attribution Models

Attribution should make your marketing efforts easier to understand, not more complex. Build your own custom attribution strategy that really benefits your marketing team and your company. Too expensive, difficult or time-consuming? Not the case - with the right help.

Data and insights experts can combine the data you already have with new ways to generate and understand first-hand customer insights. The right partner can set up unified marketing measurement that provides a holistic view of marketing campaign success and improves the customer journey to drive conversion and purchase long-term.

Viewers Logic’s media measurement tools reveal the long-term impact of your campaigns and your marketing strategy on customer acquisition. Your teams can deliver and demonstrate the true impact of advertising on revenue growth - with the right partnership.