If you want to improve results for clients, keep questioning online measurement
Andrew Bradford, VP client consulting, media, Nielsen, says we need to ensure that the industry has the knowledge to ask the right questions. Plenty is also being done by trade bodies and organisations to promote learning, but it seems clear that there is still a long way to go…
Ask people about the measurement and use of online data and you’ll get many different opinions. But the source of varying views can sometimes be difficult to pin down. For me, the key points of debate tend to fall into these eight areas:
1. Local vs international
The internet is a truly global medium and there is a demand for global consistency in how online audiences are measured. However, we should remember that advertising is still largely a local industry with local needs. It would be more accurate to describe it as a transnational industry – international companies working with distinct national practices. This difference is important and largely the reason for the creation of so many JICs in Europe and Australia.
2. Start-up innovation vs mature responsibility
The internet has been a test-bed of innovation, which has produced some fantastic insight and opportunities for targeting (eg. cookie-based; behavioural). But as revenues grow and the need to understand where the money is going increases, there will be a greater need for oversight and auditing.
3. Fragmentation vs holistic insight
Clients increasingly demand a 360-degree view of consumers, but because of the increasing complexity of online, insight is fragmenting. It’s one of the reasons behind our continuing drive to leverage the scope of our data to produce useful insights for clients, such as UKOM/NRS fusion, and TV-plus-internet fusions. Increasingly, those suppliers with niche or broad capabilities will thrive, but those stuck in the soggy middle ground will struggle to perform for increasingly demanding clients.
4. Democratisation vs professionalisation
The web has made it easier to gather information – but just because we can gather, doesn’t mean we should. A little knowledge can be a dangerous thing if we don’t consider its provenance and match it to the importance of the decisions resulting from it. It must pass the tests for being: effective, efficient, economic. It’s not always a trade off between speed versus robustness, but often that’s the dilemma. This is also a function of shorter decision-making cycles as businesses attempt to adapt to an uncertain environment, affected by simultaneous structural and cyclical pressures.
5. Internal vs industry
The internet has generated a large amount of internal data for all businesses. There has been a tendency to assume this data is perfect and that external data is less useful. External data has never been more important because control, transparency and respondent-level data is critical. Internal data is incredibly valuable, but we should not be blind to the issues around it. Device fragmentation and privacy concerns (leading to cookie deletion) are upwardly distorting trends. Site-centric datasets are ill-equipped on their own to decipher these user footprints. Likewise, panels have inherent weaknesses. Fortunately, when used together, the strengths and weaknesses of each method can be balanced out.
6. Traditional vs new culture
Some traditional media minds have struggled with the complexity of online and the need for new approaches and speed. The new culture lacks the discipline and experience of asking why and architecting data for the future. We should be actively addressing the ‘fusion’ of these subtly different knowledge bases at an industry level. It’s not acceptable for young analysts not to know what a GRP is, any more than for traditional media researchers to think of the internet as purely a medium.
7. Disclosure vs reward
Consumers’ response to the trading of their data (both descriptive and behavioural) is largely driven by their life-stage and experience online. We should be mindful of the significant differences in attitudes between groups on this matter. For some the internet represents the ability to project identity; for others it’s a black hole. These groups can create critical skews in the ability of our industry to collect representative data.
8. Qual vs Quant
Often qual is not used in online research to the degree that it should be. The complexity of the internet demands more qualitative insights to guide the quant.
So, what is the upshot of all of this? The three consistent issues that emerge are:
- the increasing need for data integration
- data provenance
- industry knowledge management
At Nielsen, we are working on solutions to these issues. Shortly, we will be launching “all-location” audience measurement, while integrating panel and site-centric data. In addition, the NRS/UKOM fusion continues, bureaus provide fusion-on-the-fly capabilities, and we’re currently working on fusing Homescan data to UKOM data.
Crucially, UKOM is allowing traditional planning (with categorisations like ‘ABC1s’) to integrate into an online buying system (with terms like ‘behavioural’) using a dataset that speaks both languages. And that’s just in my backyard of the industry.
But underpinning this is data provenance. We have to become more inquisitive, more disciplined and more transparent to ensure we, as an industry, can provide our clients with the holistic views of their consumers that they so desperately need. Online must move on from a state of passive acceptance of the first dataset that comes along, to being prepared to ask if it’s fit for purpose.
The bottom line is ensuring the industry has the knowledge to ask the right questions. For our part, Nielsen is working closely and transparently with UKOM to ensure clarity on how data is produced. Plenty is also being done by trade bodies and organisations, like MediaTel, to promote learning, but it seems clear that there is still a long way to go.