What’s next for Project Dovetail?

Harnessing the power of panel and device-based data, BARB’s Project Dovetail will witness the first stages of delivery in the coming months. Here, BARB’s CEO, Justin Sampson, explains what happens next.

Like many companies, BARB conducts an annual survey among our customers.

This time last year we were reviewing encouraging findings that endorsed both our ability to deliver a gold standard currency and also our strategy to harness device-based data with the behavioural data that come from our panel. The delivery of this strategy, widely known as Project Dovetail, involves significant reengineering so that we make the most of the complementary strengths of these different data sources.

The results from our latest customer survey are just in and there continues to be strong support for our core service. Our customers continue to give us a big tick for providing a trusted and respected gold standard.

They are also still convinced that we have the right strategy, although there’s a clear appetite for it to be delivered more quickly. Some suggest that BARB takes heed of the “good enough, fast enough” product development concept that’s become more widespread in the Internet era.

Yet BARB has to be thoughtful about the challenges of measuring online viewing. After all, internet measurement experts can be a little like economists. Give four of them the same question and you’re likely to receive four different answers. That’s before you delve into the issues of viewability, click fraud and brand safety.

Each and every data set reflects the values of the people that compile it. This means that BARB reflects the collective values of the UK television and advertising industry and we have always placed a great onus on transparency and getting the data right. It’s what our stakeholders justifiably expect of us.

Getting it right and doing it fast is clearly the best outcome, although if one of these has to take priority it is the former. After all, our audience estimates are used every day to justify the £7bn that is spent over the course of a year on the production and distribution of programme and commercial content.

Our recent customer research reinforced our belief that we have the right strategy. So how are we building confidence that our delivery is right in such a challenging market context?

Firstly, we have to ensure that we are measuring genuine behaviour. The good news on this front is that ABC (Audit Bureau of Circulations) has audited our system design and signed off that it is free of non-human and international traffic. The system is also set up to avoid any gaming of the system by the media owners that BARB measures.

Secondly, we need to know that we are getting a precise measurement of the content that is being viewed on a device. We are putting in place a precise tracking mechanism as the software code that we commissioned from Kantar Media has the benefit of tracking programme viewing activity to the second. There are no assumptions made about position in content or total time spent watching.

Another challenge is defining comparable metrics. We already have ratification from JICWEBS* for a new metric, average programme streams. This combines exposure and dwell time to provide an average duration measure. This is analogous to the average audience numbers that our customers are used to working with, although it’s a measure of devices rather than of people.

Next on the agenda is to ask JICWEBS to ratify an equivalent metric for online commercials. Delivering this will mean that advertisers and their agencies can confidently assess the combined reach and frequency of linear and online/on demand campaigns, knowing that the underlying data have been collected on a comparable basis.

The media owners are making good progress embedding the software code into their online distribution. This process has a head of steam and as the code is implemented in each TV player app, so we need to audit that this has been done in line with the system design. Again, this is being done independently by ABC.

These implementation audits are the last task to complete before we start to publish the TV Player Report. This beta report will be available as a weekly downloadable report from the BARB website and could be available early in the second quarter of the year subject to completion of the implementation audits. Not all of our existing media owner customers will to be ready to publish from the start, although we are confident that the report will become more comprehensive as the year unfolds.

Perhaps the biggest challenge is the task of integrating the different data sets. BARB is not alone in dealing with such a challenge as many companies in our industry have similar ambitions. Where we are unique is that our data fusion strategy is for the delivery of fresh, gold standard data to the whole industry each and every day. We’re yet to find another organisation that’s attempting something on a similar scale.

Given what is at stake we plan to commission two organisations, Kantar Media and Nielsen, to deliver a prototype of their proposed solutions for delivering fused data. This will take six months so we expect to review the results towards the end of this year with the view to awarding a full development contract at that point.

During this time, we will also make further progress in determining how return path data from set-top boxes can play a part in our hybrid future. Testing is underway for a development that has the potential to further transform how BARB delivers robust audience estimates in a fragmented market.

Project Dovetail is an ambitious strategy that will transform the foundations of BARB’s services. Naturally we’re pleased that our customers can clearly see the benefit of the strategy and we look forward to the first stages being delivered in the coming months.

Justin will discuss further details of Project Dovetail at MediaTel’s Video Upfronts event on 25 February. To find out more, click here.

*Joint Industry Committee for Web Standards