Sainsbury’s ex-CEO said: I’m not sure about Price Optimization
- May 8, 2015
- Type: Blog
- Featured in: LinkedIn
Earlier this year I was fortunate enough to attend the Retail Loyalty Conference hosted by BrandLoyalty. The conference had a great line-up of speakers and one of them was Justin King, the ex-CEO of Sainsbury’s, one of the largest supermarkets in the UK. I really enjoyed Justin’s talk and thought he did a great job (if you have a chance to listen to Justin speak I would encourage you to do so).
At one point during his talk he made the following statement “I’m not entirely sure about Price Optimization”. Part of me wondered if he was just being nice and actually wanted to say he didn’t believe in price optimization; a bigger part of me felt that, as far as Justin was concerned, the jury was still out on price optimization and whether it was a good solution for retailers. This comment made me ponder the question: what are the most effective mechanisms available to retailers to translate pricing strategies into pricing execution?
In the five previous pricing articles (1, 2, 3, 4, 5) we’ve talked about a variety of approaches to pricing strategy and I’m sure some of you will be asking how a retailer can actually incorporate some of these approaches into their pricing process. In addition to the challenges associated with translating pricing strategy into actual price points for millions of store/sku combinations, pricing also has to make intuitive sense to customers. In this article we’ll address the challenges of turning pricing strategy into pricing execution.
How does a retailer manage all this complexity? Lets not understate the problem: translating pricing strategy into actual price points that satisfy multiple objectives is a challenge. If an average grocer carries 40,000 items per store and they have 250 stores then there are 10 million store/SKU combinations. But that's not the end of the complexity...
If each store/SKU combination has ten viable price points and there are ten relevant data points (three competitor price points, the time of year, any relevant trends, the prices of other comparable items, any promotional activity on comparable items, the customers sensitivity to price changes, etc.) that should be taken into account when selecting the right price then you have ONE BILLION data points to consider. If a retailer considered each of these ONE BILLION data points for one minute once per year then they would need several thousand employees just to review all of these data points!
Pragmatism wins the day. So what do retailers do when confronted by an impossible task? They become pragmatic. They consider a subset of the data points, they consider a subset of the price points, they operate at the zone, region or chain level and they focus all their proactive pricing efforts on the small percent of the items that generate the bulk of the sales.
This approach just about manages to bring the problem down to a manageable level. When this is combined with some generic pricing rules that follow a market-driven approach for the KVIs with a cost-driven approach for most of the rest then pricing becomes a smaller task that can be managed around exceptions (should I follow a competitor price down? Should I match or just be within plus or minus 2%?, etc.).
So the retailer has managed to reduce the problem to a level that can be handled from an operational perspective, but what has happened to the pricing strategy, the financial objectives and the customer alignment with all these operational necessities?
Customers are not all the same. Customers that shop with a specific retailer in different stores are not all the same. If you look at customers’ sensitivities to price changes for a single item across a number of stores there are frequently vast differences. It is not unusual for customer price sensitivity for a single item to vary from twice the average in some stores down to less than half the average in other stores.
And this phenomenon is not as a result of looking at vastly different geographic locations. Even within the same metropolitan area there are oftentimes vast differences. This is a particular issue when a retailer has price zones setup by metropolitan area and has the exact same prices in stores in very different parts of a metro area.
Items within a category are not all the same. The same phenomenon can be seen within a single category in a single store. As with category roles, it is sometimes assumed that customers have a similar price sensitivity for all items within a single category in a single store. The reality is that frequently price sensitivity varies dramatically across the items in the category with some items having a price sensitivity that is much greater than the category average and other items having a price sensitivity that is much lower, even within the same store.
Operational short-cuts miss the mark with customers...and the CFO. By pricing all stores at the same level or by proactively pricing only the largest selling items retailers are missing an opportunity to consistently align prices with customer needs. Prices in some stores will be higher than customers want in some instances and lower than they need to be in other areas. The items that are priced too high for customers will lead to lower customer satisfaction and will generate lower sales volumes; the items that are priced too low could be an important source of financial returns that could be used to lower prices in other areas.
Do my prices really execute my strategy or is that an illusion? Translating a high-level pricing strategy into specific price points is challenging and it usually falls to a category manager or a pricing manager/analyst to perform the magical transformation. Oftentimes the pricing person is told they have to be competitive on a set of KVIs against a specific competitor, they have to achieve a positive consumer price perception, they have to enforce a number of pricing rules to ensure prices are intuitive to the customer and they have to grow the category by 3-5%...and they have to implement these strategies by determining the prices of thousands or millions of store/sku combinations.
The reality is that the actual prices that are deployed to stores reflect only a small portion of the overall strategy and only partially achieve some of the customer, competitor, and financial objectives. Whenever I’ve analyzed a retailer’s existing prices it is consistently clear that there is a limited connection between the underlying characteristics of the business (driven by customer needs) and the prices that are on the shelf. This is usually because the prices have not been set with an understanding of customer needs.
Squaring the circle…or octagoning the circle? Turning a circle into a square is challenging enough but effectively executing pricing strategy feels more like trying to create an octagon from the circle. So what is a retailer to do? Accept that operational simplicity should trump all other pricing objectives? Hire a huge number of people to focus relentlessly on pricing and accept the dramatic increase in costs that would come with this approach? Or embrace technology and advanced analytics in an attempt to deliver against the pricing strategy?
Can technology and analytics solve all our pricing problems? This is where technology and advanced analytics have a role to play but they have to be coupled with the retailer’s business strategy and they have to be used in an intelligent manner by experienced practitioners. Technology and advanced analytics on their own will not deliver success. I often liken pricing technology and advanced analytics to a really powerful car: a Ferrari will get you to a location faster but if you drive it in the wrong direction then you will only arrive sooner at the wrong destination!
Does price optimization actually work? When you apply an advanced analytical optimization solver in the form of price optimization it inherently uses the understanding of customer needs in the form of price sensitivities in order to set the prices. As a result, prices have a much better chance of being aligned with customer needs. But the optimization also has to execute the retailer’s strategy through the use of the right set of objectives and business rules.
I’ve seen too many situations where price optimization has been grossly mis-used but, when used correctly, price optimization will generate a balanced portfolio of prices that are not too high and not too low. In simplistic terms, with price optimization the items that customers purchase mainly based upon price receive lower prices and the items where customers purchase based on quality, convenience, availability, etc. receive a price that is commensurate with the value that customers attach to those items. This balanced portfolio of prices will ensure that customers receive the prices they want and the business achieves its strategic, customer, competitor, financial and operational objectives.
I’ve seen many retailers think that pricing software is a magic button that you push and it automagically transforms the retailer’s business but you still need a good navigator to ensure the outputs from the pricing software drive the retailer to the desired destination.
Do my prices drift out of alignment with my strategy over time? We all know that the retail marketplace is a very dynamic and fluid environment. Competitors change their prices and open new stores; manufacturers change their COGS, new items are introduced, customer needs/tastes change over time, economic circumstances change, etc.. So should a retailer react to all these changes and, if they do, does that lead the retailer to drift away from their strategy over time?
The answer to both questions is yes! With regard to the first question, the retailer should react to changing circumstances but they should do so in the knowledge of the likely impact. Following a competitor down on price when customers are not very sensitive to changes in the item’s price could be damaging to revenue and gross profits without any discernible positive impact on sales volume or consumer price perception – in this case, maybe you lower the price but don’t match the competition exactly.
What about when a manufacturer increases their cost and it drives up the retailer’s COGS? Should the retailer increase their price to the customer by the same dollar amount, the same percentage amount or should they not raise their shelf price? Once again the right approach depends on a number of factors. If the item has a high customer price sensitivity and it has a high demand transfer to a comparable/competing item then any increase in shelf price could drive customers to purchase a competing product in the category.
This outcome could be OK for the retailer if the competing item has a lower price (which could have a positive impact on price perception) and a higher gross margin…but this outcome would not be good for the manufacturer who is proposing the COGs increase as they would lose sales to a CPG competitor. Framing a conversation with the manufacturer in these terms could result in a different decision by the manufacturer. It can also ensure that the retailer is always making a fact-based, informed decision.
Over time all these small changes can lead the retailer to drift away from their original set of balanced prices that effectively implemented their pricing strategy. As a result any part of the business that experiences a lot of these types of changes should be reset at appropriate points in time. Aligning prices with customer needs is not a one-and-done effort – it needs to be performed on a periodic basis.
Self-driving cars are on the way – should we put pricing in self-drive mode too? Irrespective of how you feel about self-driving cars, I would not advocate handing your pricing over to a self-driving process. While our roads are inherently unpredictable and are subject to various dynamics I would hope that most other drivers are not out to deliberately derail you. Grocery retailing is hyper-competitive and it would be a big risk to allow pricing to run without appropriate supervision.
I believe Justin King was right to question the effectiveness of price optimization; when used incorrectly it can lead to prices spiraling out of alignment with business objectives. But pricing is too important to be left to simplistic cost-driven or market-driven tactics that are employed mainly to simplify operational processes. As with most business situations, if you put the right people in place, give them the right tools and point them in the right direction then you can achieve wonderful results; price optimization can play a central role in this process. To paraphrase a Winston Churchill quote: “price optimization is the worst form of pricing, except for all the others.”