We live in a data rich world; every interaction with an organization is recorded, codified and stored in a database ready for analysis. Private brand retail is no different in this regard, especially in the world of customer feedback. This is true for complaints, returns and reviews.
Here at S4RB our philosophy is firmly centred around connecting suppliers directly to customer feedback by sharing the incoming comments and reviews directly with the people who can actually do something about it.
As simple as this may sound, the practices and processes of feedback management need to work together to ensure the right data is available to the right people at the right time.
As part of our advisory practice, we advise and benchmark retailers on the ‘Six steps of good feedback management’:
How do you listen to your customers?
Do you understand what your customers are saying?
Can you identify what is important to look at?
Can you share it with someone to do something about it?
How do you solve it?
Do you know if you’ve made a difference?
The nature of customer feedback is subjective - both retailer and suppliers need to know they are acting on issues that have a meaningful basis for concern and cause for correction. However, in reality no one has the ability to act on every piece of feedback that comes through the door, nor should they. This is why many of our partners ask us to help in area #3 – Analysis. This involves using the data to narrow down the areas to look into in more detail.
By far the best way to understand if an increase in negative feedback has a common cause is for someone familiar with the product to read the comments and decide for themselves. Herein lies another problem: how do the experts know which comments to read and which to ignore? There is simply not enough time in the day to read them all.
The entry level technique is to simply rank by volume or by change in volume and work through the top 10%. While this is a perfectly valid approach, it carries with it the risk of missing out on the bottom 90% where movements may be more subtle, or issues may be in an early stage.
I am going to walk us through a few technical analysis techniques that can provide clues about the existence of an issue before a single comment is read. Many are based on the techniques of so-called ‘Chartists’ - technical analysts involved in financial market trading, using the ‘shape’ of data to decide if a stock is worth buying or selling, or in our case whether there is a quality issue.
Here are some areas for focus:
Complaints per million units. Quite simply the number of complaints relative to the volume sold (clearly a million can be scaled down to meet a sensible business unit). CPMU is a very quick and easy way to determine if an increase in negative feedback is disproportionate to sales or simply a result of more sales = more complaints.
The volume of complaints over a particular product or category will vary over time, but usually it can be expected to remain within a fairly predictable range, that is until an extraordinary event occurs. Setting bands based on a moving average, plus or minus a standard deviation or two can quickly identify when levels of feedback move outside of expected norms, even if volumes or movement don’t qualify for a top 10 place.
A variation on trading bands, breakout analysis looks back at the highest points over a historical time period and sets a ceiling. If and when the level of complaints breaks through this ceiling, this can be an indicator that something extraordinary is afoot.
Not all movement in quality is cliff edge extreme. Some issues develop over time and only come to the surface when they hit a level that puts them on the top 10 radar. Looking for complaint volumes or CPMU that is trending upwards for three or more periods can give a good early warning of a potential issue in the mix.
In financial services, beta is defined as the volatility of a stock relative to the market. Translating this into quality management, the movement of negative feedback of a product relative to the rest of the category, can provide nice insight into which products are more robust in the face of changing consumer preferences (clue: a low beta is preferable).
At this point it is worth noting that past performance is not indicative of future results, and clearly not all of these techniques will give definitive answers every time. As part of S4RB’s Insights module, we create collections of products based on all these techniques and more, as well as looking at the relationships between the collections.
By looking at the data through multiple lenses we give our customers the best chance of finding and addressing the key issues efficiently and effectively.
Clearly, there is no substitute for analysing the text of customer feedback and drilling into root cause (another area S4RB is leading the market on.) But perhaps by thinking like a stockbroker, any retail product manager can find issues in places they’ve never thought to look.