A customer review can produce one point of data relevant to a product. A statistically representative number of customer reviews of the same product can produce a much better quality of a single data point relevant to that product. It is a very good start, but it is still just a single point of data-CSI, CSAT, NPS, etc., depending on the methodology used to collect this data. So what is the value of this point of data? Apparently it is quite significant when used for marketing as people do pay attention to the recommendations of their “peers” or “influence-rs”. Online retailers know that conversion from visit to purchase is much higher for products that have a significant number of relatively positive reviews, and that is why they invest in collecting and managing access to these reviews. The value of this data for a product manufacturer varies from industry to industry.
A couple of weeks ago I attended a presentation by Munjal Shah, the CEO of Like.com and one of his slides really made me think.
In other words data itself is not actionable. Consider the actions a marketing product manager can take based on the data that their product ABC has a low satisfaction score. I can’t think of any other action than to learn more, i.e. to discover more data. Presumably information is created when our marketing product manager (or product marketing manager) compares ABC’s product satisfaction score with the one of a competing product, hence comparison of two points produce information, i.e. higher value.
Correlating the information produced by tracking these two data points over time with sales numbers can create knowledge – “product with an inferior reputation tends to undersell its competition by X%, when sold at competitive (i.e. similar) price”. Now, this is an actionable piece of knowledge as our MP/PM manager can attempt to discount the ABC product to stimulate sales or attempt to improve the customer’s opinion about it.
I already wrote that most CE Marketing Product managers (area of my focus) do not think that they can actively manage a reputation of a product released into the field. However this belief is not based on any wisdom, empirical knowledge, or current data. It is based on the experience of working with traditional tools in an archaic (pre-social media) market environment. The availability of customer feedback about their experience with a product, combined with modern tools, capable to extract actionable knowledge, enable organizations to create causation “pro-active product reputation management produces higher profitability than price discounting and defensive advertising”.
More From Piplzchoice
- Customers want us to listen more
- 3 simple ideas for engaging your customers to sell your products
- The CES2010 aftermath
Piplzchoice Recommends
- 10 Ways To Increase Traffic To Your Blog (Brandon Hann)
- CRM needs change: from data entry system to data aggregator (LEADS Explorer)
Social comments and analytics for this post…
This post was mentioned on Twitter by myankelovich: From Data to Wisdom – http://b2l.me/rgrb (via @piplzchoice)…
Great piece. I haven’t used the standard analytic tools for a couple years but the majority of the standard measures they present are not actionable.
Most particularly, as you’ve noted here, starting with some of the standard data, they have to be combined and compared. It’s all a matter of relativity. But there are no standards. Every company is different. The data has to be experimented with to uncover the relevant stuff.
I was once ‘let go’ by a manager who didn’t understand that this stuff isn’t self-serving. He believed that you just set up the tools and they spit stuff out. He was interested in heartbeats, which tell you nothing until you’re dead and the heartbeat stops.
[...] This post was mentioned on Twitter by Uwe Hook, Kelly Craft. Kelly Craft said: Good read. RT @uwehook: From data to wisdom http://is.gd/55YSh [...]
Sorry Paula, I just now discovered your comment. You are absolutely correct about the “meaning” of what we measure. There are two dominant approaches carrently in play:
1. Pure Discovery – We do not know what we are looking for, but we’ll see the patterns of knowledge when they emerge. I call it a serendipity approach – you are looking for the needle in a hay and find a farmer’s daughter.
2. KPI Measurements – Let us measure these things and manage our enterprise performance based on it, even though they are not related to any specific actions.
Let me suggest a 3rd or a morph of 1 and 2. The bottom line is that except for businesses that really know ALL of their channels and have fairly static products/offerings, the KPIs for 2 must constantly be adjusted by 1.
The problem is that businesses believe that the KPIs actually ‘tell’ them something meaningful. In many cases what they ‘tell’ is neither meaningful or actionable — just more of a ‘yep, it changed, but we have no idea why or what we can do about it’. If the latter, is it really worth the cost to follow/report?
I suppose the only meaningful approach to take is a “filter” approach. Monitor for signals, tune noise out and focus on reception from identifiable channels, organize/distill data into information (actionable), correlate information between the channels to create knowledge (actionable), deep analysis of multiple sets of knowledge to “figure out” the cause(s) (strategic action”. Very similar approach to mining and refinement process. Is there more fun than to noodle things out?
Then the last thing is to persist it all (including the synthesis that others can’t do — it’s not ‘obvious’ to everyone), as I recommended: http://www.fastforwardblog.com/2009/07/31/the-context-of-intent/
Yes, some of the web analytics tools provide dashboards, but this is more of a corporate synthesis and ‘reporting’ and collaborative discussion around it all: the cross-dashboard — with teases to other details (perhaps you’ve mentioned something similar that I’ve missed).
[...] need more examples of business processes where product managers have to “translate” data into “information” that suggest action. Consider the actions a marketing product manager can take based on the data [...]