Podcast: BI tools fall short with market research data
BI tools are an incredible asset but they simply can’t easily deliver the same kind of value to insights professionals that a fit-for-purpose consumer intelligence tool can for market research data.
In this "guestless" episode, our head of marketing and podcast host, Michael Howard, dives into how business intelligence (BI) tools may be fantastic for big [business] data but aren’t fit for purpose for extracting material business value out of market research data. He says they can be used, but the speed to insight is far too slow.
During the podcast, he compares Business Intelligence (BI) tools such as Power BI, Tableau, Google Data Studio, with specific market research analysis software, which he refers to as consumer intelligence (CI) tools. Many may believe that BI tools can deliver the same value as CI tools such as Harmoni. Michael notes that one thing to keep in mind through this episode is that, while it isn’t a perfect representation, BI tools are generally introspective and internally focused, whereas CI tools are outward and customer-centric - with some overlaps.
Below is a transcript of Michael's podcast, or you can listen here:
Before I get too far into the episode, I wanted to apologize if this does come across self-serving. It’s not my intention. Rather, the overall point I’m hoping that comes across loud and clear, and it’s one that our clients often feedback to us, is the importance of having fit-for-purpose tools, processes, people, and technology to help not just do your job better, but also deliver better outcomes for your clients.
So, let’s start with a bit of context…
If you’re frustrated with the tools and processes you use to analyze and report your market research data, you are not alone. Employees around the world, not just market researchers, are faced with the daily struggles of workarounds. They are given a set of tools, yet need to become “hackers” to force their tools to meet their needs. For insights functions, most data analysis and visualization tools are not built for the complexities and nuances of market research data, yet this is the reality that some of our peers face.
Frustration and poor quality are left in the wake of projects where employees are forced to operate in environments that are not fit-for-purpose – especially when needing to accomplish the sophisticated levels of analysis that market research often requires. And this is really surprising, considering how central data insights and customer centricity are to the modern business.
Now, we’re constantly having conversations with insights professionals around world. And time and time again, our frontline people get asked what’s the difference between consumer intelligence and business intelligence software.
They are most familiar with BI tools, so let's take a closer look.
Business Intelligence Tools
Some of the most popular BI tools on the market today rose to fame in the 2000s when big data was starting to take hold in the business world. Organizations were gaining access to scores and scores of data from various sources, and they needed ways to ingest and make sense of what they were pulling in.
Dashboards, the output of all the data analysis that happens under the bonnet of such BI tools, are one of the key pages of the BI playbook. Dashboards provide personalized views of what matters to both employees and those further up the hierarchy and, when designed well, are full of information that aids decision making.
They are comprehensive systems that are particularly effective in working with the likes of finance, sales, supply chain, figures, which are generally the same regardless of what industry or business you are in. With the right implementation and training, BI tools are incredibly useful in helping organizations understand trends, performance, process, and other data that is typically industry agnostic.
However, as powerful as they may be, BI tools need workarounds to cope with the inherent intricacies of advanced survey data analytics. From our 32 years of experience in the market research space, plus all the numerous market researchers who have worked for Infotools over the years, we have a solid understanding of what insight functions need to deliver value back to their business. And here are five reasons that we’ve seen as to why BI tools aren’t fit for purpose for your market research data.
BI tools can’t handle the data input.
BI tools are only as valuable as the data that goes into them. In order to maximize the return on investment of such tools, organizations need to invest in feeding them with reliable and available data. According to McKinsey, 92 percent of companies they classified as leaders in a recent survey had a process in place for tracking incomplete and inaccurate data.
Generic BI tools are typically designed for aggregate and relational data, yet struggle to process respondent-level data or multiple data inputs. If you want to bring disparate data sources together, it must be done manually before uploading or using limited data preparation functionalities. If data changes, it needs to be processed again. This is both time-consuming and error-prone.
It's essential that we look at various data sources to get a holistic view of target audiences, potentially bringing together things like survey responses, online interactions, in-store shopping data, social media posts, qualitative interviews, and more. But again, BI tools aren’t necessarily up for this level of detail. If you want to truly know your potential customers, understanding the market with the right tools is vital.
BI tools don’t allow for specific applications.
In market research, we like to look at data in a myriad of ways. When your tool doesn’t easily allow you to accomplish tasks like examining weighted data, significance testing, processing multi-response questions, or building metrics on the fly, you are at a serious disadvantage.
For example, you don’t have to had worked in market research long to know it isn't easy to obtain samples that are representative of the general population. Generally, the data will under-represent some groups and over-represent others.
If you want accurate insights, you need to weight the data. If you’re using a different type of application such as a BI tool, you will have to work really hard to apply that weighting. However, a platform that's architected for market research will do that inherently.
BI tools aren’t always user friendly.
Sharing data among wider stakeholder groups is rising in importance. Data democratization is no longer a trend but an expectation. But, when your BI tool only allows data experts with querying experience and data modelling knowledge to create meaningful analyses, many people are left out. You end up relying on just a few people who know how to do things like applying the correct function and feeding the tool the correct inputs.
In contrast, consumer intelligence tools worth your while have customer centricity built into them, and that customer being stakeholders and decision makers. Market research is a cost center for a business until it produces positive change within an organization. Therefore, usability is critical for not just getting initial sign-off for implementation, but also for ongoing value that makes the insights function look good.
BI tools don’t necessarily have built-in statistical functionalities, which are required for analyzing market research data. What this means is that BI users need statistical knowledge in order to deliver insights with a high degree of confidence. This isn’t obvious to those without a market research background, and as you don’t know what you don’t know, people are making critical business decisions based off data that is inherently skewed.
However, platforms built for market research data include statistical functionalities and other like features, which significantly opens up the tool to users outside of the insights function.
BI tools make sharing beautiful results difficult.
Visualizing and sharing market research data among stakeholders is arguably the most important part of a researcher’s job. When source data needs to be manually imported, and reports are not designed for direct client or stakeholder interaction, you run into major inefficiencies. These extra steps in producing outputs from data consume extra time and increase the potential for errors.
If you are sharing static files, like PDFs or PowerPoints that contain your insights, then people aren’t necessarily viewing the most up-to-date insights. When data visualization is a core part of your technology platform, you can easily ask questions of the research data, then quickly display answers that are easily digestible.
Purpose-built market research solutions should have reporting organized as stories that are directly connected to the data. These stories should then be easily sharable in multiple consumable formats, including online interactive dashboards with the ability to dig deeper and filter by any available variable on the fly.
Remember, the more market research analysis and visualization is shared, the more it can be used, and the more value that can be generated as a result of it.
BI tools aren’t using automation to its fullest extent.
Speed and accuracy are top demands in the market research ecosystem right now. Automation is the way to get there by completing tasks - such as bringing together multiple data sources - in minutes rather than days, meeting demands for quick turnarounds, and leaving you free to do your most valuable job of uncovering insights.
For example, intelligent market research platforms can organize, harmonize, and structure data sources to suit your needs, keeping your original data intact - including label, meta, and time information.
Finding insights should be a whole lot smarter with the right tech too. Instead of manually setting up endless crosstabs, imagine having some smart automation that can highlight significant differences or similarities among audience segments. This is about balancing the strengths of human capability with technology smarts. And when it works, it surely is a beautiful thing.
So, BI tools aren’t fit-for-purpose for market research data, but organizations are using them with workarounds.
How do you know what is one technology workaround too many?
Look, there’s no such thing as a one-size-fits-all BI tool. These software platforms simply can’t cater for all industry nuances, organization-specific processes, and employee idiosyncrasies. And I don’t think they’ll ever claim to do that either.
Humans are fascinating subject, often choosing the path of least resistance. When it comes to technology, employees insist on costly workarounds to digitize their old ways of working. In doing this, you potentially miss out on all the advances and benefits of the new platform. Furthermore, it isn’t just expensive, considering the high consultant and custom development fees; it’s risky as organizations are also left exposed if certain employees were to leave.
The other element people don’t factor into workarounds is the workload and subsequent impact on employee morale. If employees need to go to extremes to get their day-to-day jobs done, then job satisfaction rates could also be at risk of falling. And in the tight talent market we’re currently in, you don’t want to risk losing employees.
Now, if BI tools aren’t fit-for-purpose for market research data, even with workarounds, what options do we have?
Consumer Intelligence Tools
In a world of data-driven decision making and customer centricity, consumer intelligence tools are worth their weight in gold, especially ones that can provide a single voice of the customer, via integrating disparate data sources just as their Business Intelligence cousins can.
From market surveys to social listening, sales to consumer behavior, NPS to ad effectiveness, CI tools bring together a comprehensive view of your target customer and the world in which they live. The data accessible by these tools, once shaped and organized, can be drilled into in all the ways you would expect as a market research and insights professional.
It's not just a matter of being an aggregator of consumer data either, it’s the quality of that data which is just as, if not even more important.
When used well, CI tools enable you to examine customer, consumer, and other data sets – from quant to qual, structured to unstructured, that illuminate insights that are the foundation of future your competitive advantages.
Not only that, but CI tools can also help clean up consumer data so you’re feeding the organization with clean reliable insights. It’s scary to think that some samples contain up to 60% of fake, dirty, and bad responses. That is outrageous. Imagine making key business decisions based off an insight that is only 40% reliable. That sounds more like a competitive disadvantage. And that is what could well be instore for organizations who continually use tools that aren’t fit-for-purpose.
Just as BI tools with workarounds can negatively impact employee morale, businesses who invest in CI tools can positively impact employee morale, showing the business appreciates the value that insight functions offer.
One last point on CI tools, is that they’re not mutually exclusive to BI tools. In fact, some CI tools can even feed back into your organization’s BI tools, which can provide even greater value in the process of surfacing key information to stakeholders.
Being outcome focused when comparing software solutions for your insights function
If you’re on the fence about your current technology set up, it can help to work backward from your desired outcome.
A good way to think about any piece of software or technology, is what actions or outcomes you want to make or become a reality, and then work backwards to see what you need in order to achieve said outcomes or actions. If you’re making decisions based on finance or sales, then BI tools like Tableau or PowerBI would make perfect sense – and you probably have such a platform already in place.
But what about making decisions based on consumer behavior? How well can industry agnostic BI tools navigate the nuances of quantitative and qualitative data? How well can they weight and resolve sample representation issues? Software built specifically for the market research and consumer insights sector has this inherently built in (or should do anyway).
If it’s a cost-out initiative, the efficiencies that will be realized by using a fit-for-purpose, speed-to-insight technology could reduce actual cost to your business, especially when you consider consultant fees required to implement changes on large enterprise applications.
I hope this has give you some interesting food for thought on the role of business intelligence and consumer intelligence tools within the insights sector.