It’s common by now for organisations to say that they want to be more data-led. But what does that mean in practice?
Recognising the importance of data is a great start, but implemented incorrectly, that can mean just adding extra sources of confusion. You can get bogged down in statistics and fast facts, looking for a needle in a haystack, but not even knowing which field that haystack is in.
In reality, what organisations should be aiming for is to be led not by data but by insight. To continue with the agricultural metaphor, think of data as raw grain and insight as flour. It’s more refined, useable, valuable and nutritional, ready to fuel your business for the years to come.
So how do you become insight-driven? How does a good research agency help you turn data into insight?
Goals win games
Every research project should start with a conversation and commitment to establish the client’s problem. Without fully understanding and empathising with that pain point, and without keeping that in mind at every stage of the process, there’s no hope of generating a sensible solution.
Some research projects start with assumptions to challenge or validate. Let’s say a client suspects that a recent price rise or the emergence of a new competitor with funky colourways has contributed to falling sales. At other times, the agency be working with a blank slate. Sales have dropped but the client has no idea why. Both are fine, but either way the researcher must maintain an open mind. It might be that your competitor has vultured some of your customers, but for a totally different reason you didn’t envisage. It could be comfort not colourways. Sustainability not price. That means a good research agency needs to, literally, keep its options open, investigating not just the factors that you assume are in play but other crucial or linked topics too.
A research agency shouldn’t rule things out at this early stage. And if a client has tunnel vision at this point, the researcher should be helping them see the light. Tease and test and funnel down. Keeping eyes and ears open now will help the agency stay open minded when it comes to interpreting the results and recommending a solution.
Finding the right approach
Research is replete with intimidating terminology, and it doesn’t help that the industry has often been guilty of pushing pre-packaged services rather than taking a genuinely consultative approach. The client doesn’t necessarily know if they need quantitative or qualitative research, let alone what the best mix of techniques and services might be. It’s a research agency’s job to figure that out, based on understanding a client’s objectives and challenges and to define a programme that offers the best chance of delivering a solution, keeping in mind industry best practices in terms of data storage, cleaning, and checking.
The fieldwork stage is no time for corner cutting. As a researcher, it’s imperative to remember that your client wouldn’t put a sub-par product on the shelves or work with a substandard supplier. So when you’re representing them as an insights agency, that means you also have a responsibility, if required in areas like survey distribution or translation, to outsources certain elements only to the highest quality partners.
Once it comes to designing a questionnaire or putting together scripts, asking the right questions is a product of expertise. The agency needs to know the client and their industry inside out, as well as be impeccably versed in research best practices. And no, asking Chat GPT for a list of questions to then plug into SurveyMonkey is not a good substitute. That’s like taking a plastic shopping bag on a hike instead of a good rucksack. They both technically hold stuff, but one is going to hold up a whole lot better.
Draw on other data sets for context
Sometimes, you need a little more than traditional analysis to help turn data into insight. One option is to lean on additional proprietary data that you may have access to. Whether it’s sell-through / sales data or participation or activity frequency statistics, qualitative or quantitative data sets can all provide vital context to the analyses you’ve conducted.
At other times, the additional context you need for your results needs to be sought outside your or your client’s organisation. Happily there’s a wealth of data available in the public domain that can add depth to your findings. Whether it’s economic trends, stock prices, employment data, consumer or retail price indexes, weather data, facility usage, publicly available participation numbers, there’s often extra context to be found if you’re willing to go searching for it.
Lean on leaders
Nobody better understands how to make decisions than people who have experience making them. Our team includes people who have worked in senior leadership positions including CMO, MD, CEO. In other words, people who are used to taking decisions using data and able to draw on the types of questions and analyses that matter. Their presence ensures that insights and recommendations are reasonable, feasible, and targeted to help address the client’s challenges. It’s the kind of thing they’d want access to in order to make a difference in that senior role.
Be curious
This advice will be familiar to fans of Ted Lasso (or Walt Whitman who Ted took it from), but perhaps the best mental prompt to foster insights and not data is simply to “be curious”.
Be willing to go beyond, to not take things at face value. To rip up your crosstabs and start again. To investigate the internal or external data that will give you that crucial extra nugget of understanding. To distinguish between correlation and causation, to think deeply about connecting factors, limitations, and possibilities. To find the story in the statistics, and then put it under stress to make sure it stands up to scrutiny.
Great researchers are both archaeologists and sociologists. They’re able to find the small little nuggets of broken pottery hidden within a large data set, to think carefully, based on a close understanding of industry and culture what that could mean. But they’ve also got to understand how humans think and act, individually and in groups.
Being bullish without spouting bullshit
There are two major failings that research agencies make when delivering results. The first is not being bold enough. Data becomes a safety blanket, and your reporting ends up not going any further than convoluted charts and graphs. Your summaries are statistics rather than stories. Failings of curiosity and of confidence result in an absence of recommendations. The client ends up in much the same place they were before, but with more distracting data.
Your recommendations should be built on the data analysed (and so turned into insight) of course, as well as on institutional and personal expertise. You’ve likely dealt with many more of these kinds of datasets than your clients. Your team has the expertise to discuss and game out options. That should give you confidence to step up and explain what, in your view, the client should do next.
On the flip side, many a business has fallen for gaudy looking results and implemented changes only to later realise that sample size or other factors mean that what they thought was a clear steer was actually statistically insignificant or represented correlation not causation.
The client has spent the money and so it’s tempting to feel like you have to force a solution, but it’s imperative to avoid getting carried away with inconclusive findings. Sometimes its ok to say more study is needed. Extra research might be cheaper for the client in the long-run than taking a costly trip down the wrong path.