Digital transformation. Anyone who works in or adjacent to a marketing function will have surely heard this dreaded buzzword. I certainly know I have — a lot. Shorthand for adopting technological replacements to manual or analogue processes, digital transformation shares the same status as miracle cures. It can improve culture, business efficiency, customer experiences and adaptability. In short, in the digital age — why shouldn’t everything be digital?
Well, as you may have guessed from the title of this piece, the answer is not straightforward. First, let’s look at the statistics. Any kind of transformation is hard. In fact, a McKinsey & Company analysis found that only 30% of business transformations succeed. The success rate of digital transformation is even lower — half that figure again. And, an addition 7% of businesses report that any improvements are not sustained over a significant period.
The same report suggests that both digitally savvy and traditionally analogue industries struggle, with success rates ranging from 4% to 25%. Considering how broad in scope such changes are, this should come as little surprise. Making widespread changes that affect multiple departments, operating processes and existing systems is never a painless process. But, even so, digital transformation appears to be a harsh gamble; high risk for high reward.
In the insights industry, I believe this leaves us with two key questions. First, should research teams and agencies attempt a digital transformation? And second, should our work be considered in the context of a digital-first environment?
Research Agencies and Teams
Every insight team — regardless of agency or client environment — will need to, at some point, make a decision on whether to pursue digital transformation as a way to improve efficiency and external cohesion. However, in our line of work, transformation of this nature hits a natural and early roadblock. While digital-first tools, methodologies, reporting processes and communication may be an initially attractive prospect, there is a looming question about how it impacts the ability to understand and interpret customer behaviour.
People make decisions in an analogue manner. While it is true that digital footprints are increasing, technology has yet to solve all the frustrations researchers have — even with digitally native methodologies. Take online focus groups for example. The largest time-sink with this method of qualitative data collection is not running the groups, but rather the analysis and reporting. Why? Because to find and present the most relevant snippets requires re-watching what can turn into hours of footage. Automatic transcription and search software can speed up this process immensely. But even so, accuracy rates remain low. Research from Descript estimates that most widely available AI-based transcription services have a 16%-37% error rate.
For researchers, that’s a wide margin of error. It’s one that makes it difficult to justify fully relying on digital processing. An element of human intervention is still required. And this is a pattern that repeats across multiple domains. Sub-par survey completes can be spotted by AI and machine learning tools to a certain extent, but human review can still improve upon it. Technology based image tagging can speed up visual processing, but it is still far from perfect.
These trends make it easy to question whether digital transformation can be effectively achieved in the insights industry with our current technology. I’d contest that the greatest need right now is not, in fact, for transformation — but integration. It’s impossible to ignore the obvious advantages that technology offers the insights industry. But equally, whether it is because of the analogue nature of human decision making or the very practical limitations of current automation — true transformation is hard to achieve. So we should focus on making the connection between the two smoother, more efficient and more aligned to our needs. In some instances that may be an 80/20 split, in others closer to 50/50. Either way, we should automate what we can, invest human resource in what we can’t and seek out the best way to marry these two approaches.
Digital First Environments
The second question I think it’s worth considering doesn’t so much relate to the internal processes of insight teams, but rather their output. As an increasing number of organisations adopt a digital-first approach, how do we transform or integrate research outputs in a way that still works.
I think there are a few ways to answer this. But perhaps most important is to again consider the benefits of both the digital and analogue. Digital distribution of insights can make important information:
- Easier to search for, access and share
- More instant and readily available
- Collaborative, and more discussion-oriented
- Simpler to organise, sort and store
Conversely, analogue modes of data presentation can be more impactful, more human and more persuasive. Additionally, particularly creative methods of analogue reporting can better interrupt, stand out and may therefore be remembered for longer. The greatest case for analogue reporting is made by a now famous comic strip. In the left panel (representing the past), a cartoon figure ignores a full mailbox to read a single email. In the right panel (representing the now), the same figure ignores a full inbox to read a single letter. Standing out matters.
Once again, I believe the best approach for the insights industry to take is to blend these two methods of reporting. That’s not to say every project and tracker should have both a digital and physical report. Instead, the measured approach is to develop a process that identifies the weight of decisions any project will influence. Using this measure, research teams and agencies can then determine which factor is most important; ease of access or memorability.
Integration vs Transformation
I realise to some this distinction may feel like an argument of semantics, but there’s an important distinction to make. In process, in culture, in methodology and in reporting, the insights industry is given value by both the automated and the manual. To transform into a solely digital field is both practically challenging and leaves behind a significant amount of value we draw from the very human side of the profession.
If value is derived from both, we should not aspire to forsake one approach over the other. We should become better at identifying where each is best applied, and finding the cohesive glue that joins the two.
This article was originally published on the FlexMR Insight Blog and can be accessed here.