In the article below, RetailWire talks with Weather Expert Paul Walsh on customer behavior, and what retailers are doing to balance things out.
“What almost always happens when you get a big event like this is you really see three phases from a retailer perspective,” Walsh told Retail Dive. “Actually, it’s really from a consumer perspective, and the retailers’ job is to be ready for their customers’ needs.”
Retailers don’t lose or gain sales as weather events hit, unless their supply chain is not prepared for the push ahead of the event. What this signals is that retailers are getting out ahead of these events using weather data, and suppliers are being increasingly asked to help anticipate supply chain strains ahead of the curve as well.
“The weather ate my homework” is no longer going to be a viable excuse in the coming months and years. There will be an expectation that a supplier network will need to be out ahead of this. It won’t matter if you are a top tier supplier maximizing your category, a mid-tier supplier scratching and clawing for their business every day, or a small supplier just trying to keep up. The forecasting models are becoming more and more accurate. Not having your products attributed to weather sensitivity will be a hindrance.
It’s also important to consider “just weather data” may not be enough. Feature engineering both the weather data to tailor it to your products and your products attributes may be required to extract enough insights as to what weather patterns and events are effecting your sales. Make sure you have people on board that can do this with your data, and it may take some experimentation, since no two product assortments are the same. Having the data at the same level as your geography dimension also helps take some of the burden off of your analytics platform. Its a lot of data, and most of it comes at the weather station level.
Have I mentioned yet that we can do this for you? We can provide weather data at the zip code level, and we have experts on staff that can help with your feature engineering, as well as your reporting and data mining.
You can read the whole article here
Please contact us today to see how we can help you with your BI weather challenges.
After reading the online article – “3 big data types that will increase your ROI” – it struck me as something that CPG companies who sell directly to large retailers sometimes limit their view to the data the retailer provides back to them. Make no mistake – that data is essential to a profitable partnership with the retailer. But with innovations in analytics, big data and cloud computing richer, more robust, datasets can provide a deeper story than POS data alone.
Two of these data types have relevance to how CPG companies interact with their retailer.
By now, most companies have some sort of social platform or presence in place. A Facebook page to share content, a LinkedIn group to network with prospects, or perhaps a Pinterest page to showcase products. Many marketers are hyper-focused on measuring the number of followers, retweets or shares their pages and content are getting. This is certainly a good thing to measure and engaging with your customers and prospects on social platforms is a must. However, there are huge opportunities within social media that can lead to even bigger wins.
Knowing this information can increase the value of the analytics you share with your buyer.
Hard to Find Data
Numerous, hard-to-find data assets such as these can be sourced from the big data universe through a data-as-a-service solutions provider. What’s powerful about using a data-as-a-service solutions provider, like Vortisieze, is that that you don’t need to implement a big data system or hire data scientists to start accessing this data. The insights have already been mined and sourced, and can be integrated directly into your database.
For CPG, weather data has been difficult to incorporate into traditional analytics, however, Vortisieze has incorporated this alongside POS, inventory and demographic datasets, just to name a few.
With traditional relational database management systems (RDBMS) adding unstructured datasets is problematic, time consuming and expensive. With Vortisieze as your data-as-a-service solutions provider adding new unstructured data is fast, low-cost and actionable with your retailer.
There are lots of ways to analyze your POS sales against weather, but it might be good to first determine if there is even a correlation between the two. If you’ve got weather data and POS data in your DWH, you can achieve this relatively quickly by using a weighted trend graph in MicroStrategy’s Visual Insights. As you can see in the screenshot above, we created a data set with Walmart week, some POS metrics, and some weather metrics – the one we are interested in for this set of products is Average Temperature. With WM Week as the x-Axis and POS sales as the y-Axis, we then have the ability to color the line by Average Temperature, and we weight the line with precipitation. As the line moves from red to blue the temps get cooler, and as the line gets thicker we see more total precipitation across the country for that week.
At a high level, this graph doesn’t give us anything we can action directly on. However, it does seem that as the country cools off we sell substantially more product. With a little more work we can add some drill downs to state or store and look at temps down to store level to see if each store exhibits the same behavior or if they seem unaffected.
Not all products have a correlation to weather, but these seem to, so next post we will dive deeper into how weather seems to affect sales, and we will try to add some more interesting metrics to help identify outliers or patterns we can predict. Feel free to comment or send suggestions on how you would like to see this dashboard evolve to email@example.com and we will see if we can incorporate your suggestions into the dashboard.