I like the IIeX conferences. They’re a great way to spend a couple of days seeing innovations in the industry, learning what different brands are up to and connecting with startups.
(I always try to get on the agenda so I get a free ticket … but if you don’t fancy being on stage, it’s still good value.)
Here are my five key take-outs from the last couple of days in Amsterdam.
Privacy regulation isn’t crushing innovation – it’s spurring it
Privacy is thankfully becoming a big deal.
When GDPR came in last year, it felt like an endless compliance nightmare for many companies. And some legal departments continue to frustrate marketing and insight teams with a default NO setting for any request; but most have adapted, and the long run aims of the regulations are to be welcomed.
There have been some positive short-term impacts as well: dodgy players in the programmatic ad space have left the EU; most sensible businesses now model their worldwide policies on the standard; and the innovation winter the surveillance capitalists predicted hasn’t materialised.
In fact, many companies have spotted the commercial opportunities in building more privacy-first solutions to research and analytics challenges. They’re ahead of the curve.
Powrofyou is a privacy-first personal data exchange, which allows users to share their smartphone data anonymously and receive payment in return.
Its Qualify product makes its technology available to survey panels via SDK. With users’ permission, it identifies digital behaviour from the preceding 90 days (apps used, sites visited, search terms entered) – and uses this for targeting or pre-screening of surveys.
It’s a smart use of behavioural data in an exchange model, and as one of the Innovation Competition entires, it beat a field of 20-odd other startups to win the $20k prize. Congrats.
Both are building blockchain-enabled solutions to connect different data sources (survey panels, DMPs etc) in privacy-first marketplaces for personal data.
You can see other examples – not from IIEX – along with some principles for working with behavioural data in this article.
The best insight teams are fully data source agnostic
People should stop hand-wringing about how research needs to move on from surveys and focus groups.
That ship has sailed.
Most decent insight teams – and many research agencies – now use dozens of different sources to build their evidence.
Elaine Rodrigo, Chief Strategy & Insights Officer at Danone, showed how her team used machine learning to combing text and image data in social media with other data and primary research.
A panel debate on the organisational impact of combing research and data analytics functions featured leaders from a wide variety of companies.
Turner’s Pedro Cosa talked about combining different data sources from panels, surveys and digital measurement into a holistic KPI framework for executives.
Mail Metro Media’s Becky Hillcoat showed how connecting web analytics and primary research built a strong commercial case for premium ad inventory.
Nestlé’s Liz Boffey explained how the global insight team now has access to over 100 different data sources and tools.
It’s not about primary research any more; it’s about getting data from the right place for your brief.
AI has fully landed in research & analytics
This time next year, we’ll all be SO OVER Artificial Intelligence. It will be 2020’s blockchain, we’ll all be eye-rolling at it.
In the meantime, AI is a useful catch-all to help galvanise the research industry around machine learning. It is already delivering step change improvements throughout the industry (read the Guide to AI for Research & Analytics here if you haven’t already).
In Amsterdam, there were some great examples of AI-enhanced approaches.
(And, to be fair, very little lipstick-on-a-pig re-badging of old stuff with a new ‘AI’ label.)
Klydo is an innovation research platform powered by NLP. Tens of thousands of product descriptions are scraped from e-commerce sites like Amazon; these are then decoded with NLP to map proposition territories and identify opportunities – in white space or trending areas. Unilever is a customer.
Market Logic Software showed off impressive work with Coca Cola’s S&I Connect (it’s Research & Knowledge Management platform).
It combines cognitive search capabilities with Natural Language Generation: users can ask the system, “what do we know about X?” – and it will synthesise the most appropriate answers from validated research projects (there are over 95,000 reports in the database), social media intelligence and other sources.
Connie Zhang of Abbott presented work with Quid, whose NLP platform analyses and visualises the content of datasets (including patents, company releases and millions of news articles and blog posts). It was able to define and map its competitive position in the “Life Changing Health Tech” sector and use the insights for communications planning.
And Black Swan’s Richard Maryniak talked through the use of social data analytics to map, rank and predict upcoming consumer trends much earlier than primary research – with case studies from PepsiCo, Danone and McDonald’s.
It’s all about embedding insight
This means different things to different people. To me, it means two things.
Putting insight capabilities directly into the workflows of stakeholder teams: giving them the tools to gather their own data, analyse it and act on it.
In some organisations, this happens without insight teams even knowing about it: brand, digital and e-commerce teams just crack on and buy their own tools. In others, the insight team works closely with stakeholders to design the process, build the guard-rails and choose a supplier. Sometimes you can’t let people mark their own homework.
Activating insights created by the research team. This is all about influence and driving commercial decisions – better storytelling, creative outputs, the marketing of insight.
(You can read more about embedding insight here.)
At IIEX, there were good examples of both.
In the first camp, Fastuna is a relatively new addition to the research automation category. Focusing on simplicity and ease of use, its growth in its home market (Russia) has all been down to getting non-researchers in marketing teams using its tools as an integral part of their creative and planning toolbox.
In the same vein, KnowledgeHound presented startup lessons for established corporate teams; their analytics platform makes survey research data available to users in all departments, not just the insight team.
And if activating insight is the raison d’être behind Coca Cola’s S&I Connect platform, it’s also the driver for ‘Naked Insights’ – showcased by Coke’s Begoña Fafian.
Working with marketing agency Keen as Mustard, the insight team treated its research output as a creative campaign – testing different executions, using a mix of formats (video, infographics, newsletters), and being unafraid to take a provocative stance.
Hence ‘Naked Insights’ as the campaign tagline.
‘Build or Buy’ is a false dichotomy – for agencies and for clients
Both businesses have transformed their insight operating models, data sources and technology platforms over the last few years.
Unilever’s seemingly aggressive mission – double the value, half the cost, half the time – has seen it review more than 3,500 startups over 5 years and pilot some 800 new tools.
And Kantar has a $150m Capex budget for R&D – much of which it ploughs into technology development.
But for both businesses, it’s no longer a simple question of building or buying tech.
After initially partnering with Zappi for research automation, Kantar is now launching its own Marketplace solution; but it also has a range of tech partnerships including Affectiva for copy testing, CINT for sampling and has recently licensed IRI’s Liquid Data platform for Worldpanel reporting.
Unilever runs pilots with startups; backs those it regards as promising to help them scale rapidly, and even takes equity stakes in those it wants to push particularly fast.
These may be two of the biggest players in their respective fields, but their approach to tech development holds lessons for insight teams and agencies of all sizes: there are many creative models for building tech capability – develop in-house, collaborate, invest, revenue-share, joint venture – and you don’t need to be constrained by straight build-or-buy business cases.
I’ve not done any kind of justice to all the innovation and thinking on display over the last couple of days, but hopefully this gives you a flavor of what went down.
If you’re attending the IIeX Austin event and want to write it up for this site, drop me a line.
This blog was originally posted on Insight Platforms.