Marketing attribution is a way of understanding the impact of different channels and touch points on customer decisions and spending. It allows businesses to allocate responsibility for sales (and therefore income) to particular channels or campaigns, and see which are most effective in driving sales and profits, and which give the best return on investment.

Marketing attribution is not new, although you might think it from some of the hype. It has been around for a considerable time. Early models tended to focus on attributing sales to a single touch point, perhaps the first or last point of contact with the company (on the grounds that without the first contact, nothing would have happened, or that the last point was the one that finally drove the sale). Others tried to share out the responsibility equally between all the touch points. None of these, however, was very sophisticated. Broadly speaking, they do not take into account customer behaviour, or consider differences in influence between channels.

Modelling has become much more accurate and effective, mostly as a result of applying more sophisticated analytical techniques, and particularly algorithmic modelling. The best approaches tend to combine bottom-up information, such as analysis of customer behaviour and channel usage, with top-down information, such as changes in spending on particular channels. The new modelling techniques also recognise the differential influence of multiple channels, and can take into account each touch point’s influence on later progress. This is a huge step forward in understanding the customer journey, and in turn has major benefits for organisations.

Attribution modelling driving organisational benefits

Assessing return on investment of marketing spending has always been difficult. The improved accuracy of attribution modelling has now, however, made this significantly easier, and provides a number of additional benefits, as well as the obvious. So what are these benefits?

Detailed information on channel and campaign performance

First, and perhaps most obviously, attribution modelling allows companies to see and assess the broad performance of particular channels and campaigns. The information available is much more detailed than was possible before. This means that companies can assess the effects of each step in the process on particular downstream actions, such as searching for more information or comparing prices. It is also possible to see whether that particular touch point has a direct connection to sales. The level of detail means that contributions to revenue or sales can be allocated much more accurately. In practice this has huge benefits for marketers because campaigns and channels that are not performing can be reduced or stopped altogether. The resources used can be switched to more effective options, saving money, and also improving marketing performance more generally.

Clear view of the whole purchasing process

Attribution modelling also provides a much clearer view of the whole purchasing process from start to finish. Data is collected at every touch point and used to model the process more accurately, including on an individual level. This gives a much better understanding of customer behaviour, and therefore means that marketing can be targeted more effectively, both in terms of time and the precise offer made. With more information available, it is even possible to see which keywords are most effective.

Bringing online and offline together

Attribution modelling can be used for both digital and real-world channels. Bringing data from online and offline worlds together provides a much fuller picture of customer behaviour, and at a high level of detail. This, in turn, informs decisions about channel mix, including offline channels, and ensures that they fully reflect customer preferences.

Making data-driven decisions with confidence

It is this capacity to inform decisions that is arguably the most important aspect of using attribution modelling for marketing. Companies that use these techniques can make decisions about marketing spending in the sure and certain knowledge of what actions will be most effective in driving sales. There is no more guesswork involved, only informed, data-driven decision making.

This is a huge benefit. Taking the guesswork out of marketing makes it much more effective. It also ensures that decisions can be made more quickly because there is no more arguing for pet projects on the grounds of instinct and anecdote. Marketing attribution modelling supports the metrics that matter to businesses, such as profits, turnover and return on investment, and links marketing spending directly to these.

Marketing attribution modelling is only likely to get more impressive as more data becomes available on customer activity, preferences and behaviour, and as modelling techniques improve. It will be particularly interesting to see the contribution made by artificial intelligence in this field, but I look forward to progress all round.