As firms start to understand more about how customers interact with their advertising, Andrew Hood foresees a need for a more robust way to allocate payments between different elements of marketing campaigns.

Companies often work to ‘Cost Per Acquisition’ (CPA) targets to benchmark how much they have to invest in a particular online marketing campaign to win one customer.

But visitors interact with multiple digital marketing campaigns from email and search, to search and affiliates in the lead up to acquisition. 

Naïvely attributing acquisition to a single campaign can mean paying two, three or even more times over in reality. Only by pulling all online marketing activity together and understanding the interactions and reactions, can the real cost of acquiring new customers be established.

The concept of an integrated and inter-dependent marketing mix is nothing new. The offline world has always sought after that mystical optimum mix of TV, PR, radio, outdoor, direct mail (and so on) to maximise return on marketing budgets.

The challenge has always been finding the right formula for splitting budgets in a context where measuring the impact of specific campaigns is often, at best ‘finger in air’, spawning a whole research and econometrics industry to try to make sense of it.


But online is different. Every digital marketing campaign can be tracked and measured in terms of sales, revenue and leads generated.

The success of every click from a search engine can be quantified, the value generated by every email campaign discovered and the pounds and pence racked up against the required investment.

And perhaps the relative ease of tracking individual campaigns online has resulted in a convenient but misplaced simplification: the assumption that every online marketing element operates in a vacuum.

That simplification becomes dangerous when it comes to measuring CPA and in turn, overall return on investment.

Consumers research products online over time (particularly in industries like travel and financial services where shopping around is commonplace) and a sale may well be a result of more than one marketing interaction and critically, more than one marketing cost.

For example, a visitor may initially click through to your site from a sponsored link in the search engines, have a browse, then recognising your brand a day later, click through on one of your banner adverts and make a purchase.

Traditional ROI reporting would attribute that sale to the banner click, ignoring the earlier search visit. Arguably the banner click only happened on the back of the earlier search click and this dependence is lost in evaluating the performance of the campaigns.


Some of these correlations can be pretty stark. Recent research by Lynchpin Analytics across its client base found on average that over 50% of online sales driven by banner advertising were reliant on earlier visits from other channels such as search and email.

Even more subtle trends, such as 5% of affiliate-sourced customers returning via PPC to buy later, demonstrate an overlap that can have an impact on budgets (especially with rising CPC in paid search).

The concept of an integrated digital marketing mix thus becomes critical to making sound judgements on CPA and forging strategies to maximise ROI.

If there is a high correlation between two campaigns or channels in delivering sales, dropping one because the other appears to be more efficient when measured in a vacuum can easily kill the bottom line.

Equally, a campaign that looks to be performing brilliantly may actually be reliant on the presence of other channels.


Quantifying these correlations can be a complex issue. A current topic of debate is how to split the value of online sales across the sequence of marketing events leading up to them.

Is every click across the research-buy cycle as important? Should the first click carry extra credit for introducing the prospect to the website in the first place? Is the last click prior to purchase the deal-maker? Do branded keywords attract more or less value?

As companies start to develop a more detailed understanding of how customers interact with their marketing, online and offline, in terms of generating awareness, capturing interest and motivating purchase, robust models will no doubt emerge for splitting the costs.

Meanwhile, perhaps a simpler approach is more realistic. If we can see the high-level of correlation between campaigns and channels that generate results, we know which channels need to be aligned in terms of marketing investment to maximise the impact of the synergy.

Equally, if there is a strong overlap between two expensive campaigns, we know to raise our CPA expectations and feed that back into budgeting.

The transparency and accountability of online marketing has fuelled its growth with quantifiable CPA and ROI.

As the market matures and becomes increasingly competitive and multi-channel, the challenge will be to understand the reactions and interactions in digital marketing needed to forge integrated strategies for engaging customers across the purchase cycle.

Those that only measure the effects of individual campaigns in a vacuum, risk missing a trick in terms of a realistic view of their customer acquisition costs and suffering diminishing returns from their engagement online.

Andrew Hood is the managing director of Lynchpin Analytics.