Brands want more insights from retail POS data
Glenn Taylor
Through a special arrangement, presented here for discussion is a summary of a current article from the Retail TouchPoints website.
While brand manufacturers continue to invest in POS analytics and gather relevant customer data from retailer partners, 77 percent still believe they can be doing more with the data to improve their business, according to a report from Askuity.
Beyond internal decision-making, the data is used in buying and planning discussions with stores.
“On average, 65 percent of brands are actively using POS data today, and that’s actually quite encouraging and exciting,” said Victor Coscarella, Askuity’s head of marketing, during a recent webinar. “If we were to do this 10 years ago, we imagine that this number would be far lower. It could even be below 50 percent.”
The report is based on 343 respondents from more than 150 vendors that sell in one or more national retailers.
The report still found 28 percent of brands are not making use of POS data despite many major chains providing the data free of charge. The primary challenges include having an undefined strategy for using POS data, poor data quality, and a lack of analytics expertise or technology to manage the data effectively.
Managing data can be cumbersome. Nearly half of the respondents reported spending at least four hours per week managing or essentially formatting the data, even before any top-level analysis.
Silos are an issue. When data is managed and distributed by a single power user within the company, brands are far less likely to bring data insights into buyer meetings (41 percent), versus brands that provide self-service access to POS data for all team members (67 percent). Forty-four percent are not providing their field sales teams with access to real-time POS and/or inventory data on a mobile device.
The top ways brands are looking to extract more value from POS data:
- Better tracking of new product launches (59 percent);
- Improving promotional effectiveness (56 percent);
- Improving forecasting to reduce safety stock levels (55 percent);
- Optimizing merchandise assortment (55 percent);
- Strengthening relationships with retail buyers (51 percent) ;
- Improving on-shelf availability (39 percent).
Discussion Questions
DISCUSSION QUESTIONS: What are the primary hurdles preventing brands from fully capitalizing on the availability of POS data? In what ways may stores be inhibiting collaboration around sales and inventory data?
Poll
BrainTrust
Mark Ryski
Founder, CEO & Author, HeadCount Corporation
Ron Margulis
Managing Director, RAM Communications
I’m having some difficulty believing some of this data. A third of CPG and grocery manufacturers are not using POS data? Yes, there may be a bunch of very small companies who can’t afford to buy syndicated data. It’s unclear if these companies would be able to make use of it with buyers even if they had it. The biggest hurdle is believing that the data contains insights — it’s just data. The insights come when you ask questions and have someone (or hire someone) smart enough to use the data to answer the question.
Effective data analysis still remains an aspiration for many organizations — it should be a core competency. As a company that specializes in analyzing retailer data, I can relate to all of the challenges described in the article. Data is often poorly managed and full of impurities, and retailers often lack the internal expertise and resources to fully extract the value/insight that should come from the mountain of data they collect. POS and inventory data are critical elements, but these are only two of many, including store traffic data, labor data, marketing data and more. Often these data sources need to be combined to find the critical insights — which is even more challenging than analyzing any one data set, like POS data alone.
In the age of Big Data and powerful analytics, retailers and brands need to apply the resources necessary to be effective. Collecting data is easy, extracting the insights and then applying them in a way that leads to better business decisions is significantly harder. Retailers that effectively manage and analyze their data have a significant advantage over those who do not.
Data is always better than no data. I have seen POS extraction complexities if it involves integration or aggregation difficulty in pulling data across more than one POS type, but all of it is the cost of doing business today — data is a must.
Sourcing and analyzing the data is step one. The most effective retailers (and brands) are using channels to act on that data. Those that can communicate with consumers on a personalized basis or inject offers back into the POS are a step ahead of the game.
Access to data is the easy part. Knowing what to look for that can significantly impact both the business AND benefit the shopper remains a challenge. Often times the retailer is only focused on its own broader success and frankly doesn’t really care about the individual brands. The brands, on the other hand, have everything to gain by measuring their success. The brands are interested in better understanding how they are doing compared to their competitors in a particular category and retailers are reluctant to share the broader category data.
Insights from data cut both ways and, as such, the insights shed light on issues that many folks don’t want to know! These insights can show that initiatives designed, or favored, by agencies or executives that have cost time and money simply don’t work.
A true 360-degree view of the buying process simply isn’t available — it’s hard to get in every customer’s head as she or he considers what to purchase when and where. So algorithms based on some major assumptions are developed and available data is put through the analytics engine to deliver knowledge that can be acted on. If the analysis is good, or just plain lucky, the resulting demand creation messaging is effective in steering shoppers to products and services via select channels. If not, well, the markdown rack gets some more inventory.
The biggest challenge facing manufacturers is the channelization of the buying process. It’s hard to get a true picture of which marketing and merchandising tactics are working when you don’t have eyes on what’s selling (or not selling) where. Even when companies do have access to t-log data, there is so much noise and delay that the information delivered often isn’t actionable. There are tech vendors who have great solutions for the full view of the customer, but ultimately it’s up to the manufacturers and retailers to implement systems that can quickly share analytics, not just data, so the customer can be appropriately marketed to.
Collaboration and data sharing between retailers and brands is a relatively new phenomenon. It is only in recent years that retailers have been willing to share POS data with brands, so it’s not surprising to see brands having issues around data formats and access. As retailers get more comfortable sharing data and brands become more familiar with the POS data structure, the two sides will realize more benefits from collaboration.
First of all, Mark and Adrian are absolutely correct in their comments. I’ve said for a long time that data does not equal insights and insights alone are not enough. Only actionable recommendations move the needle. If the manufacturer and retailer cannot agree on an actionable recommendation then nothing happens. All the data in the world does not change this simple fact. Many of our clients do use POS data, but mostly along with many other data sources including syndicated, panel, shopper and other data sets that allow for a richer set of facts, to arrive at insights that hopefully will lead to action.
The challenge with getting from data to actionable insights and recommendations is the data prep. This is the biggest time hog of an analyst’s day. Data is messy and there is a good deal of work involved to get it ready for reporting. Adding disparate data sources, which is critical for a seat at the table with the most sophisticated retailers, makes it even harder and more time consuming. There are tools, from us and others, that do a really good job of significantly reducing this data prep time from days and weeks to minutes.
The reason this is important is that if you don’t give the analyst tools to clean the data, you will burn them out. Another reason is speed. One of the most challenging hurdles most manufacturers need to overcome is speed to insights. If you don’t have a process to prep the data and get the reporting out to the field with deliverables that are simple to understand and simple to present — fast — you will be at a huge disadvantage to those that do. It can be done, but organizations need to make a commitment to this and not just give it lip service.
This is all about two key challenges: 1) True collaboration and, 2) Knowing which technologies to leverage. This perennial challenge has existed since the first product was scanned in a supermarket back in 1974 … maybe even prior to that. The time has come for all retailers of all sizes to connect with all of their CPG suppliers and define a strategy for data sharing. That’s easier said than done, though. If there is currently not a strategy in place, start with your largest suppliers and determine what can be captured and shared.
From there, the next major step is to investigate the best technologies available to achieve your objectives. The good news is that there have been huge advancements in only the past several months. With 80% of all data being literally invisible to most retailers’ systems, you need to capture the right data first.
I’m not in agreement with the stats listed in the findings in this article, however, retailers and CPGs can definitely improve from where they are currently.
Two comments here:
First, recognizing 20 years ago that POS data was devoid of any information whatsoever about what was happening in the store, but only what happened at the checkout, led to the invention of PathTracker(TM) with a whole new suite of metrics. The problem with that was we were just adding an enhancement to POS data, that this whole discussion is about. However, I have been spending more than a decade trying to figure out how to globalize (potentially every store in the world) our growing data engine. Cost and simplicity are HUGE challenges that we are making GREAT progress on.
Second, I cite a Charles Schwab, (long ago steel magnate,) method for management of numbers in chapter 5 of my second edition of “Inside the Mind of the Shopper.” Basically, the problem is producing numbers which are OBVIOUS as to what actions need to be taken. It should NOT require someone with advanced data analytics training and experience to make use of reports. The rule is, “the fewer numbers, the better.” And every number’s utility to the enterprise should be obvious. We are making serious progress on this basic requisite for survival in the oceans of big data!
What are we expecting brands will learn from the data that can’t be learned any other way AND offers a significant impact on their business? Data, more often than not, is the realm of tweaks — offering only small improvements.
Up against this, remember that data is expensive. There’s the cost of gathering good, useful, and valid data. There’s the organizational risk of the data being breached. There’s the cost of analysis. And there’s a huge cost searching hard enough to find something significant.
Even when retailers supply it for free, there has to be meaning found in it. And that’s a far harder problem. So I’m not surprised that brand usage of data is less than we were told to expect (by the big data advocates).
So here’s what I recommend: Always approach data with caution. And remember the key term in research being “actionable.” Does the data offer any insights that you can act on to significantly improve your profitability? Sometimes there will be. But be careful, it will be actionable far less often than we’re told by those selling data services.