Do retailers need better business intelligence tools or a better analytic strategy?

Do retailers need better business intelligence tools or a better analytic strategy?

Most retailers still use Excel for data analysis. Looking back to the days of Business Objects, Cognos and Brio, many tried business intelligence software with the promise of doing something more valuable than what is possible with a spreadsheet — make better decisions.

Despite this, shelfware conversations were rampant in those days.

It’s fascinating how even today most companies report struggles with data analysis. Managers say:

  • It’s too hard;
  • The right data is not available;
  • IT has to be involved, and they are intractable;
  • Aggregated data just tells you what happened — not what’s going to happen or what you should do about it;
  • We can’t take action on our analysis.

Human nature may be at work here. Expectations escalate in relationship to how well a manager leverages analytics to improve performance. Analytic excellence also depends a lot on information architecture. So maybe IT is at fault?

What’s missing here is what is now proven: data-driven excellence starts at the top, with a CEO who values and has at least a conversational knowledge of modern analytics, such as artificial intelligence.

At a conference earlier this year, I started a talk with a quote from a ZDNet article — “There’s gold in them thar databases.”

It read: “Although many companies have successfully implemented data warehouses — massive databases containing large volumes of historical data for analysis and reuse — many more have struggled to do more with that data than run basic reports using simple tools.” 

The quote was published in 2003. How little has changed.

Within the last 20 years, companies such as Walmart, Kroger, P&G, Unilever and Starbucks have gone down the analytics path with decent results compared to most in their industries.

The capabilities offered by certain software today mirrors how these leaders take insights and activate them into business processes at scale.

As managers evaluate new options for analytics, they should not make the same mistakes as in the past. Instead, any evaluation should connect to a CEO-led initiative to improve the business via analytics and focus first on use cases that support the business strategy.

Like digital transformation, analytic excellence is a journey as opposed to an end state. It is all about continuous innovation.

BrainTrust

"Actionable insights are the difference between being around in five years or getting left behind. "

Jeff Weidauer

President, SSR Retail LLC


"The reality is that the analytical approach thus adopted is always retrospective – even if you are attempting to use Excel to build out predictive models..."

Oliver Guy

Global Industry Architect, Microsoft Retail


"If ever there was an area of the retail enterprise that cried out for top-down leadership, it’s analytics."

Nikki Baird

VP of Strategy, Aptos


Discussion Questions

DISCUSSION QUESTIONS: Are retailers making improvements in their approaches to data analysis or are they still stuck using spreadsheets to manage their businesses? What do you think is holding retailers back from achieving the benefits of advanced analytics so widely reported by industry leaders?

Poll

21 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Mark Ryski
Noble Member
4 years ago

Retailers are becoming better at analyzing data and applying insights, but there’s still a long way to go. From my experience, the challenge retailers have is not in capturing data – they all have more data than they know what to do with – it’s in extracting the critical insights and applying the insights to decision making. And while I agree that the CEO and leadership need to drive the data strategy, the real value is in getting simple, practical insights to the field managers and then training and encouraging them to use the insights to make better decisions in their day-to-day work. When it comes to analytics in retailing, this is and remains a work-in-progress.

Jeff Weidauer
Jeff Weidauer
Member
4 years ago

I suspect that if you polled retailers, you’d find that the majority still use spreadsheets to run their businesses. There are several barriers to making the transition to more sophisticated tools, but all could be solved if the CEO truly understood what was at stake, what was possible, and made advanced analytics a priority. Data mining might not be as sexy as a new robot or a mobile app, but actionable insights are the difference between being around in five years or getting left behind. This is where company leadership has to — well — lead.

Nikki Baird
Active Member
4 years ago

I still see too many retailers who approach analytics as a free-for-all. Business functions get to define their own metrics, often based on what they can get out of reporting off of operational systems, with IT along for the ride, trying to reconcile different ways of defining the same metric across multiple areas of the business. And when operational reporting isn’t enough, because there is no common language for sharing results across the business, they turn to spreadsheets as the expedient approach to get what they need, rather than wrestle with IT to do it right in the first place.

If ever there was an area of the retail enterprise that cried out for top-down leadership, it’s analytics. Letting each department define how they want to track their business area for themselves is like letting the accounting department say they want to operate in Spanish, the merchandising department in Chinese, the store ops group in Russian, etc. If that sounds ridiculous, just try asking each of those groups to define “sales.”

Cathy Hotka
Trusted Member
Reply to  Nikki Baird
4 years ago

Completely true. It is a free-for-all. Perhaps the more fundamental question is whether retail companies have a shared vision of where they want to go; I have yet to meet anyone who says they do.

Anne Howe
Anne Howe
Member
4 years ago

Many retailers are still asking old questions like “what was purchased?” instead of addressing “what are customers not buying here and why?” Simple question, but for a grocery retailer the impact is huge if I’m going across the street to Costco for better grade meats and $3 savings per bottle on my wine. It’s time to re-think what’s being asked.

Rob Gallo
Rob Gallo
4 years ago

Progress is being made but the challenges are myriad. It’s not too hard despite the complaints. However, you do have to take the time to create an analytics strategy. That would include an exhaustive assessment of what data you have, what data you really need and a roadmap to fill any gaps. This is a leadership directive that should be cross-functional in nature. Strategies should also include a resource plan to find the talent that can turn data into actionable insights, changes to the organizational structure (if necessary) and a change management plan to ensure that everything is implemented properly.

Ray Riley
Member
4 years ago

It comes down to the quality of data on the input side, and the retail point-of-sale in many of these cases is to blame. It’s a crowded ecosystem of middleware and aggregation tools, but until the source data gets it right the fanciest tools or strategies are always an uphill battle.

Oliver Guy
Member
4 years ago

Excel is still the lowest common denominator when it comes to reporting or indeed any more advanced form of analytics. The ability to easily bring multiple pieces of data from disparate systems together in one place still eludes most businesses – no matter what the industry.

The reality is that the analytical approach thus adopted is always retrospective – even if you are attempting to use Excel to build out predictive models (which in itself is not a good idea). The retrospective approach means organizations are missing out on opportunities. A vast majority of insights from data are perishable – in that you can do something useful with it, but only if you can capture it, combine it with other data and act on the insight it gives you immediately.

The biggest barrier to this is data silos that exist within organizations – which only seem to increase. Organizations need to be able to connect multiple data streams and sources together – irrespective of technology, heritage or vendor and be able to monitor them in real-time for data events – or non-events – that have significance. Being able to use predictive models within this approach is also important but ultimately it leads to the ability to automate response to specific events – increasing efficiency and improving customer experience.

Ricardo Belmar
Active Member
4 years ago

Retailers certainly do not lack data. What they lack more and more is the expertise and skillset needed to derive insights from that data. It’s not just a leadership question. Yes, analytics leadership does start at the top, but retail CEOs need to not only lead in pushing for more and better analytics but also define the objectives. Too many retailers approach data analytics in multiple directions through internal silos but this needs to change, or the value of the insights will be unfocused and fail to move the business forward. One of the other aspects holding retailers back has more to do with acquiring skilled labor – people with the requisite data science skills needed to make this activity successful. This is a very competitive workforce, not just for retailers, but across industries and retailers will need to up their game if they’re going to attract the best talent in this space.

Ralph Jacobson
Member
4 years ago

Great thoughts, Gib! Adoption by retailers of tools other than Excel has been slow. However, we are seeing the innovators take advantage of deep analytics to drive tangible results. Those that aren’t are literally being left to play catch up.

Ryan Mathews
Trusted Member
4 years ago

Part of this issue is frankly generational. If you “made your bones” using spreadsheets, that’s likely what you will continue to use. As to the larger question, let’s not confuse tools with the nature of the problem that they are being used on and/or the skills of the toolmaker. The issue here is creating a new vision of what retail could be, not adopting new tools for old practices.

Paco Underhill
Paco Underhill
4 years ago

Gathering data in 2019 is easy. Figuring out what to do with it is the tough part. Part of the problem is that the modern tech tools tend only to process what’s there – they can’t see what’s not there and should be. Retail is the reflection of social change – yes there are biological constants, but much is in flux. Spreadsheets and databases are tools that are only as useful as the people and brains using them. American retail historically is about birth, life, death, and compost. As any farmer can tell you a certain amount of compost is good – it frees up room and people for the next birth…

Jeff Sward
Noble Member
4 years ago

I’m not sure we need to malign spreadsheets in order to elevate the conversation about analytics. I think it is more of an ego and learning problem. There probably are new tools that can help create enhanced context and then present the information in more easily digestible formats. The person on the receiving end still has to be a willing student. Learning and changing from within is powerful — and too rare. Would better analytics have saved the many retailers that have gone bankrupt in the past couple of years? Really?

Doug Garnett
Active Member
4 years ago

Doing anything worthwhile with data doesn’t start with the data nor does it start with “analysts” — it starts with a smart team who spends considerable time determining what the important questions are.

Unfortunately, this is a rare find. I recently saw it stated the the problem in data science is that data scientists are more interested in doing clever tricks than seriously understanding the business in order to do what’s important. Here I remember the quote passed along by W. Edwards Deming: Applied statistics is 5% statistics and 95% application understanding.

All that said, expectations are far too high for data. Truth is, there is no data strategy that is going to be able to drive business. Success comes only through strategies which rely on good data. Data is secondary — not primary.

(For credentials: I have an MA – Applied Math from University of California (SD) and have spent years analyzing retail data with clients.)

Peter Charness
Trusted Member
4 years ago

Sometimes the challenge isn’t in the BI tools that are presenting the answers … the challenge is in asking the right questions. Whether the results show up in Excel or another presentation tool misses the point.

Casey Golden
Member
4 years ago

The fashion industry runs on spreadsheets. For many reasons, software fails in these organizations. I believe more niche analytic tools will enter the marketplace to address specific business processes and encompass the trade secrets that have kept business managers in Excel. A one size fits all or highly customizable system is not going to get the adoption needed to replace spreadsheets. People’s livelihood relies on their numbers; asking someone to change to a solution that is not a perfect fit is unrealistic. Every department runs on different driving KPIs. The greatest advantage will be contextual consumer data that is able to be sliced and diced across divisions to provide value — not just ads.

Craig Sundstrom
Craig Sundstrom
Noble Member
4 years ago

A tool is just that — a tool. If the user has no idea what to do with it, nothing much will happen. And of course, as we’ve said here on RW (at least once a week), execution matters. If I had a choice between a brilliant, even highly competent CEO with Excel and some dullard with the latest analytical software, I’d go with the former.

Camille P. Schuster, PhD.
Member
4 years ago

The tools are only as good as the people using them. What are the tools supposed to do? Which tools do it best? What training is necessary to use the tools to make relevant decisions? This is another example of pursuing two directions and making them intersect.

Peter Luff
4 years ago

I agree it is highly important for the CEO to value the key KPIs, such as conversion rate, drawn from the data warehouse. However, it is critical that this key data is easily available, without all the retailers staff needing to become a database analyst. Approaches and solutions need to gather the data and more importantly to supply insights driving specific direct actions for the team. This will allow the retail team to do what they are good at — concentrate on shoppers and provide great insights through clear direction derived from the data.

Phil Chesterton
Phil Chesterton
4 years ago

There are a number of stages that a retailer needs to go through to become truly proficient in data analytics.

Firstly they need a source of clean, reliable, relevant data.

Once this (far from simple) step has been achieved, the retailer can move to providing key, timely information to those associated who need it. I’m talking fairly straightforward reporting here.

After this, machine learning and AI can be employed to take the retailer to the goal of not just analyzing what is happening but what will, or needs to, happen.

The number of retailers that aren’t even at stage one in this sequence never ceases to astound me.

Get the basics right and don’t try to run before you can walk.

Gary Read
4 years ago

The barrier to advanced analytics for retailers may not lie in the actual analysis of the data, but in the accessibility of the external data retailers need to inform their strategies. Any retail data analysis should include information pulled from both internal (pricing/inventory/sales/etc.) and external sources (competitor pricing/inventory/sales/etc.), but pulling data from external sources is not an easy task. Many retailers have invested large amounts of resources and manpower to create a reliable flow of this external data to be analyzed. This manual data harvesting requires standardization and organization of the external data, which can be a costly and time-consuming process — reducing the amount of time and resources available to actually analyze the data.

What retailers need is a data-focused leadership team, not one merely focused on the analytics process, as well as a strong Chief Data Officer to guide data collection and initiatives. An automated tool that can clean and format external data as it is collected eliminates most, if not all, of the need for manual standardization of the data. Some web data integration tools can standardize, analyze, and visualize the data in one simple step, and actually integrate the data insights directly into existing business systems — making the data immediately actionable for advanced analytics vs. requiring standardization prior to analysis.