This month we’re chatting with Jon Hackenbroch, Head of the Bureau Reporting Service, about insight reporting and automation.
Firstly, tell us a bit about the Bureau Reporting Service and what it is you do?
Bureau Reporting is a service where you can outsource your reporting needs. We are a team of advanced users of the E-Tabs Enterprise software, all very experienced in all types of market research projects. These can range from the one-off small projects to large continuous/repeating projects. We pride ourselves on being an extension to your team, we often advise and provide added value to projects.
We work with both research agencies and end-user firms and enjoy any challenge thrown at us. If you are spending time making charts/reports we can definitely help.
So when should someone consider utilising report automation?
The best projects will be anything that has some form of repetition; this can be in the number of reports or even a single report with multiple sections, like multi-country / brand / Segment studies etc. We are able to handle report variations well, so definitely something you should consider when your brand awareness reports have changing competitor brands or question differences.
Quick wins returning on investment could be monthly KPI studies, often small in size but frequent and with tight turnaround times. You can automate these at fairly low cost and remove any of the charting hassles from reporting.
Any project with a tight turnaround time would also be great for automation. You can build reports in advance using interim or dummy tables, all the setup and testing can be done in advance, with the final runs being produced as soon as the data is available.
You’ve been involved in market research reporting for many years now, what would you say have been some of the biggest changes you’ve witnessed?
Dare I say PowerPoint reports haven’t changed a great deal, we went from large decks to more condensed reports; but I am starting to see very large decks again.
There has definitely been a move to making reports look more modern looking / unique. Less boring charts, more images or icons.
Reporting seems to be less pressured, timelines are better managed which means less last minute rush. Planning has improved and therefore expectations have become more realistic.
Drill down reports have become more common, where a small PowerPoint template is used to produce multiple filter cuts of the data. This could be the PowerPoint way of the dashboard ability to filter on different variables.
I have seen a better format of data tables, it is rare we see text tables. Almost all clients will be working with excel tables now.
You obviously see a lot of research decks pass by your desk, are there any report design mistakes you see often that perhaps we, as researchers, should try to avoid making?
I have an issue with overuse of significance tests. Comparing current time to many years ago will often be insignificant and won’t make for better research. Usually, so many things have changed in that time that even comparing is questionable. Showing multiple layers of stat testing on single values.
Lack of planning or thinking through results vs design. So often the design of reports look great from the designers, but when you put live data in it looks very different. Test templates with more realistic data. Think through requests that sound ok in design but in practice will look very different.
Another thing would be flat-line line charts, reporting continuous KPI project that shows no change, or limited change. This doesn’t tend to add much value to the reporting.
You and your team have been automating a part of the research process for over 15 years, what are your thoughts on the recent industry focus on research automation?
I am excited about the talk of automation, this is now coupled with the move towards AI/IA. The challenge we have always had with automation was the lack of standards in the industry, between MR teams, and even within the same teams. Hopefully, the dream of AI/IA will encourage the move to standards and then better automation. I was surprised the result in the GRIT report said 57% of research agencies and 51% of research buyers were automating charting & infographics. This included companies using and under consideration, so there is still a fair way to go, but the recent hype will help find the standards needed for automation/AI. Eventually, the gap between collection and reporting/insight will disappear.
Do you have any other tips for us in regards to report automation?
Automation needs to be well planned, don’t get distracted by the one “impossibility”, focus on the bulk. I always train my staff to focus on delivering 99 out of 100 slides, rather than just the one complex slide. 80% automation, for example, is already a huge benefit.
Before automating you need to be clear on what it is you are trying to achieve, make sure you and your team are aligned with what you want. We often get conflicting instructions, or team members will not know enough about the project – we can become the source to historical project requirements, explaining why something was done. Document decisions and reasons; annotate templates with specs and notes. This is important for internal management and continuity.
If you would like to speak to Jon about your reporting, please fill in the form below and someone will be in touch.