Why did UKOM change supplier?

It is important that UKOM is progressive and that it delivers to the UK industry the best possible data for online audience measurement.

More than two years of broad industry consultation and a rigorous RFP process outlined that significant advances were required because the media landscape has shifted, with an increased demand for a multi-media perspective. With a single source panel at its core that delivers respondent level data critical for such a perspective, Ipsos iris was unanimously selected by the UKOM board members as best placed to serve the current and future needs of online audience measurement in the UK.

Why did UKOM select Ipsos iris over others?

While all solutions that were presented to UKOM took broadly similar approaches with a hybrid methodology combining a people panel with census, or tagging, data from websites and mobile apps, Ipsos iris delivers a single source panel of 10,000 individuals and around 25,000 desktop, tablet and smartphone devices. This makes the Ipsos iris data more representative, less modelled and therefore more robust than previous methodologies which involved multiple panels. Ipsos iris was built ground up, taking into account the needs of the industry and clients when consider the user functionality and reporting of data.

Why is the Ipsos iris panel better?

A single source panel is interoperable with other data sets and will allow greater integration of UKOM endorsed data with a number of proprietary data sources.

Not only is the Ipsos iris panel demographically representative of the UK online population, the data can now also be reported by region (as per BARB TV regions) and also by ethnicity.

There is some daily audience data available in Ipsos iris that has not been available before and the inclusion of more than 40 media consumption and lifestyle questions in the panel questionnaire allows  further, more granular, elements of audience categorisation.

Can I compare data across the previous and current suppliers?

The two data sets should not be compared. No part of the UKOM RFP required that the supplier of the new contract generate online audience data that was directly comparable to the previous supplier’s data. A new supplier, a new panel, new tagging and new methodologies and modelling mean that no consistent ‘exchange rate’ can be applied at any level between the old and new data sets. There are strong precedents [link to Richard Marks piece] for taking this approach when a currency has a fundamental change.

Are the available universes measured the same across the two data sets?

No. The current universe measured by Ipsos iris is c47m people aged 15+ across all devices on a single panel. The previous supplier modelled its data from multiple panels with a universe of 6+ on desktop and 13+ on smartphones and tablets: a combined available universe of c.52m people. Ipsos is currently measuring a universe that is c.10% smaller than the predecessor.

The change in universe will have some effect on the change in data sets but that factor cannot simply be applied to the new data set. Different behaviour by different age groups, across different categories of content and across different devices, along with the new panel and new methodologies means that no such simple factors should be applied.

So, was the old data wrong?

No. Through selecting, defining and governing Ipsos iris, UKOM has moved from one estimate of online audiences to a different way of measuring those audiences that gets us closer to a ‘truth’, but neither supplier’s measurement was or is the absolute truth.

According to the new data my audience has gone down/up from previous data. Why?

The audience to your website is unlikely to have changed dramatically. With a new data set in Ipsos iris, the way UKOM now measures your audience has changed.

You should check with Ipsos that your site’s ‘dictionary’ is accurate and fully represents your web and app properties.  If you have deployed Ipsos iris tags you might check the level of tagging being used in reporting.

What has changed around measuring or reporting audiences delivered by social media/distributed content?

In principle very little, although as the methodology is different to previous UKOM suppliers the output numbers may differ slightly. Ipsos and UKOM respect that some platforms reserve the right to refuse tagging directly within their platforms. Ipsos are working directly with these companies to onboard the relevant data by other means, typically a server to server data exchange, in the same manner as previous suppliers. Examples of this approach are Apple News and Google YouTube.

Where publishers deploy Ipsos iris tagging to their own content, when it is viewed within other platforms, e.g. Facebook Instant Articles, the technology is such that this ‘distributed content’ can also be measured and credited to the source publisher.

Why does the change in methodology not have a consistent impact on publishers’ numbers?

It is not only possible but probable that publishers in the same sector may see different changes between Ipsos iris audience data and that of the previous supplier. Among a range of factors, a new panel, new methodology and updated publisher organisation and website structures with a different level of tags in place, mean that no simple exchange rate mechanism that can be applied across the two data sets.

Some publishers also fully embrace distributing their content across other platforms whilst others are more selective, Ipsos iris coverage of platform distribution channels may also influence the reported audience volumes.

Can I compare category data across old and new supplier data?

No. Changes in both publisher structures, tagging and an updated category taxonomy means that you should not compare Ipsos iris category data with that of the previous supplier.

With the new data set some numbers fluctuate across months. Why? 

Online audience behaviour is dynamic and UKOM requires that Ipsos keep the Ipsos iris capabilities as up to date as possible, so that the data can be as robust as it can be at any given time. This includes continually expanding the census or tagging of sites and apps, and the fine-tuning of the modelling process. The flexibility that UKOM requires of Ipsos iris to continually evolve means that reported audience data will also evolve.

The last provider measured from age 6+ on PCs. Why does the new supplier only measure 15+?

The single source panel nature of the Ipsos iris solution means that the recruitment of panelists as young as six for passive on-device measurement is unworkable although the current solution may be adapted in the second half of 2021 to measure from the age of 12+. UKOM and Ipsos are consulting with the industry on developing possible solutions for online audience measurement of children under the age of 12.

For international businesses is there a similar measurement in other countries that I can compare?

UK Online Measurement (UKOM) is responsible for the industry standard of online audience measurement in the UK only. Ipsos may operate similar measurement solutions elsewhere globally but it is not UKOM’s responsibility to ensure consistency across countries.


See also this FAQs page with quick links to more detailed content

« Back to News & Views