What is good incremental reach?

Phil Sumner, Nielsen’s UK media director, uses new research insights to answer a question often asked by the industry but has, until now, largely gone unanswered…

What is good incremental reach? 2%, 5%, 10%, 15%? It’s a question I’m often asked by advertisers and agencies, and one I’ve heard markedly different answers to. But, in reality, the data just hasn’t been available to support a definitive answer across various scenarios.

Knowing a campaign’s incremental reach is important because it allows planners to allocate marketing spend in the most effective and efficient way.

Over the years, there have been several stand-alone studies from Nielsen and other research houses that have tackled the question and, in general, most results seem to suggest an average level of incremental reach in the region of 1%-5%. While these pieces of research are valid and offer considerable data on specific campaigns, it has been difficult to move beyond individual experiments towards something using a consistent methodology and providing a definitive industry-accepted view as to ‘what is good incremental reach?’

Taking a step back, the question itself is too simplistic.

Firstly, I’ve met many advertisers who see incremental reach as wastage, so it’s important to establish whether incremental reach is the desired goal for the campaign. If a campaign’s objectives are to drive frequency – thus re-enforcing a TV message – then the goal may well be to minimise incremental reach.

Secondly, I think the question should be: ‘what would be a good level of incremental reach based on the size of my campaign?’ And taking this one step further: ‘what would be a good level of incremental reach based on the size of my campaign and the balance of TV and Online media weight used?’

Evidently, this question is more of a mouthful, but we need to understand these factors before we answer the question.

In late 2013, Nielsen launched Cross-Platform Campaigns Ratings in the UK. The purpose of this solution – often dubbed XCR – is to help advertisers better understand the relationship between TV and online media and to accurately establish a measure for incremental reach in a consistent manner.

By Q1 2014, we had enough data to start mining and extracting learnings, and these have significantly progressed our thinking around what is good incremental reach for any given campaign.

So, what have we learnt?

Firstly, incremental reach differs dramatically. The highest level observed so far is over 14%, and the lowest, less than 0.5%. Clearly there is a relationship between TV weight and incremental reach, e.g. a campaign with a high TV weight will naturally have less chance of achieving it. Our earliest models during beta testing were based purely on TV weight, and although they showed a strong relationship between TV weight and incremental reach, this relationship wasn’t strong enough to be the only defining factor.

The next iteration of the model ingested the TV impressions for the campaign and weighted the TV reach based on the proportion of the overall campaign impressions that were displayed on TV.

For example, where a campaign has a 60% TV reach, and the TV element of the campaign impressions were 66% of the overall campaign impressions, the weighted TV reach would be calculated as 39.6% (i.e. 60% x 66%).

Fig.1 shows our findings to date. Taking the above worked example, our findings suggest that a campaign with 39.6% weighted TV Reach should expect to generate around 4% incremental reach.

Likewise, a campaign with 20% weighted TV reach should expect to generate around 10% incremental reach.

For the first time, we’re able to build an accurate incremental reach curve that will allow advertisers and media planners to predict what they should expect based on campaign size and proportion.

At this early stage, we’re looking at the broad picture and we’re only just lifting the lid on this relationship. What’s exciting is the ability to produce incremental reach curves for specific demographics, media types and segments.

Equally, it will be possible to isolate single publishers, such as Facebook, and establish an incremental reach curve against a regular TV buy. This will come once we have a critical mass of data to support a robust analysis.

In the meantime, there are methodologies available that allow advertisers to better understand the incremental reach relationship for an individual campaign, such as Media Inventory Optimisation (ShareShift) scenario generators.

This type of research takes a single cross-media campaign, which has aired, and allows advertisers to run several reallocation scenarios, pinpointing the one which would’ve been optimal for the campaign objectives.

Furthermore, new solutions, such as Specific Media’s TV Audience Segments, give advertisers the ability to buy advertising that targets light TV viewers, therefore maximising the opportunity for incremental reach as the campaign is served.

The theory is that these targeting solutions shift the curve in Fig.1 to the right and offer advertisers a cost effective, real-time way of achieving incremental reach.

Clearly, we’re at the start of a long journey. What we have established so far should be treated as broad learnings, as each campaign will be different based on its individual media choices. But this does give advertisers an early benchmark and it does answer the wider question of whether their campaign received ‘good’ incremental reach or not.

Feel free to contact Phil to continue the debate, or leave a comment below.