I’ve been deluged with calls and emails from suppliers whose products address the challenge a reader posed in my previous article. So much so that it has taken a few weeks to pore through the suggestions, solutions, and pitches I received.
If you have already forgotten the initial query, here is the problem we’ve been hoping to shed some light on:
I work as a visual merchandising analyst at a major apparel manufacturer and retailer. We are trying to optimize our merchandising of individual product placements and thereby increase conversion.
We have a couple of thousand products in the catalog, and 180 prominent placements that we use to feature products (home page, category pages, etc.). We can measure individual products sales and draw some inferences about placement, but sales are not directly connected to site traffic or placement.
So, to really track 180 products, their placements (both of which rotate at least weekly), and their resulting sales from those placements etc., is very manual and daunting — not to mention trying to draw meaningful conclusions and devise a merchandising strategy that will put the right product in the right place at the right time.
Can you think of a good strategy to link placement, product, and sales together to be able to measure return and devise a strategy to improve results? Or, in lieu of having a tool that will tie these elements together, how would you suggest going about improving our decisions about placing product? (Today merchandisers are working mainly on instinct and replacing products that don’t perform after a week.)
The responses spanned quite a spectrum, indicating that there are a lot of different ways to attack the problem. Though that’s not surprising, I sympathize with the dilemma of those of you seeking easy answers… none of this is easy or straightforward yet. Not that the solutions don’t hold promise, but determining which approach makes the most sense in each situation is a daunting task. Just evaluating all the ideas that came to me was enough to cause a profound feeling of being overwhelmed! And that’s before any issues of moving into the final selection and implementation phases.
But as pioneers in the online measurement space, we have to expect such hardships, and the process of choosing an approach and selecting solutions is a process worth investing time in. So if you are brave enough to strike out into uncharted territory, we’ll look at some of the solutions that readers suggested in response to the challenge.
You’ll be seeing a lot of good product information here over the next several weeks, but today I’ll start by passing along some insights that take a more conceptual perspective, offering a way to think about the measurement challenge.
Dave Reiner from NetGenesis took the request for a short (!) elevator pitch literally, and resisting the urge to offer a hard sell on NetGenesis products instead offered some thoughts on a measurement approach. It’s so succinct, that I’ve included his remarks here as sent (with only a slight amount of editing): You found that sales aren’t directly connected to site traffic or placement. Not surprising, since this is at heart a multidimensional analysis problem and there are some critical dimensions and e-metrics missing.
To start, I’d suggest three e-metrics to measure:
- Visit counts to pages with product placements
- Click-through rates on the product placements themselves
- Conversion rates (sales) once the product click-through has happened
A minimum set of dimensions I’d track would be these:
- Product (e.g., banded-sleeve mesh polo shirts)
- Product category (e.g., casual tops)
- Target shopper segment (e.g., women)
- Placement type (e.g., top-level category page)
- Placement location (e.g., top middle of page)
- Source of shopper (e.g., came to site from a specific banner ad elsewhere)
- Time of day (e.g., 8-10 p.m.)
- Previous areas of site visited (e.g., end-of-summer specials page)
- Subsequent areas of site visited (e.g., casual shoe category page)
- Product price, shopper segment, and shopper value (e.g., profitability code) — if these are available from your sales database
Analyzing these measures with respect to this set of dimensions will help you understand what combinations of factors drive purchase behavior at your site. It will allow you to take direct action in segmenting your online shoppers, targeting their interests, optimizing product placement, purchasing ads, and tuning your site for improved navigation.
I’d love to talk more about this key area of strategy, but we’ve arrived at your floor… Thanks, Dave, for the overview of factors to watch. Next week, we’ll start with some specific solutions to consider.
Whether you are just getting started in the e-metrics arena or like some of my most responsive readers have been grappling with these issues for a long time, there is an event on the horizon that you ought to know about.
Jim Sterne, fellow ClickZ writer, a regular contributor to this column and a co-author of the E-Metrics white paper we pointed you to in the spring, is hosting an E-Metrics Summit in Santa Barbara in early November.
Emily Ma, product director of Tencent’s advertising platform products department, was a keynote speaker at ClickZ Live Shanghai where she discussed the ... read more
In today's multichannel world how can marketers use data to ensure the experience a customer receives is relevant to them?
The terms that customers type into your site search function can help you to gain an understanding of user behaviour and can be used to optimise ... read more