Given this is a somewhat-hypothetical historical scenario, let’s make a couple assumptions to simplify:
1) We’re going to “imagine” we’re in the beginning of 2008. At the time, eBay charged sellers for the “Gallery” feature.
2) Let’s assume any eBay revenue figures or eBay Gross Merchandise Volume (dollar value of goods transacted) are “current” time period (any time period you want to use within the last year or so).
3) Feel free to make other simplifying assumptions as-needed
We are trying to decide whether to make the “Gallery” listing upgrade free to our sellers on eBay.com (US business only). eBay generates a large portion of its US marketplaces revenue figure from charging sellers additional listing upgrade fees. We’ll say the annualized revenues from just the gallery fee are $150MM for the US business (eBay.com). While there are obvious disadvantages to making such a large revenue stream go away, there are some advantages, as well.
Here’s an explanation of the Gallery feature, in case you’re unfamiliar with it. When a buyer searches on a keyword (for example, “laptop”), in the search results that are returned there will be a seller-provided image to the left of each listing that has paid for the Gallery feature. If the seller has not provided an image, then instead we display a tiny green cartoon-like camera. See the attached ppt slide for examples of what listings with the gallery feature look like compared to those without. (Pay no attention to the fact that the sample images are from the German site — I just wanted to provide a visual for you). Gallery had roughly 65% adoption in the US at the beginning of 2008, so for an “average” search you might see 65% of the listings with images and 35% with tiny green cameras, though this adoption level would vary (for example, queries on high-value goods vs. low-value goods might see higher or lower levels of gallery).
We had a hypothesis that if we made the gallery feature free, that we could increase buyer activity, so we decided to run a test. Here’s some background on the test:
We directed one group of visitors to the site to an experience that was exactly like the site default experience, where 65% of listings had gallery adoption (showed an photograph next to the listing). We directed a second group of visitors to an experience where we displayed “nearly” 100% of listings with a gallery image in search results. We were able to do this because for the vast majority of listings (say, over 98%), the seller supplies at least one image or there is a stock photograph for listings that sellers have tagged to our product catalog.
We ran this test experience with two groups for a week and measured buyer activity on the site for these groups. We captured bids+bins, purchases, and the “GMV” (Gross Merchandise Volume — the total dollar value that buyers paid the sellers) for those purchases. I’ve attached the results from the test in the excel workbook. In the “Default site experience” group, visitors to the site would see green cameras next to each listing that did not pay for gallery. In the “All Gallery experience” group, all listings would have a photograph next to it in search results, though we still had a record of whether the seller had actually paid for the gallery feature. Since this paid/unpaid gallery determination is based on the listing, we were able to divide the buyer activity that is directed to the different listing type. You’ll notice that the data for each group is broken down by the dimension of “gallery feature paid” and “gallery feature not paid”, and also provided in total.
You have two objectives:
1) Interpret the test data and draw a conclusion on the following question — were we able to increase buyer activity by simulating the “free gallery” experience? Are you able to learn anything else about the buyer experience that is interesting? Please summarize your findings with an executive audience in mind (fairly high-level conclusions) and use tables, charts, text as you see appropriate.
2) Your goal is to develop a recommendation as to whether we should offer the gallery feature for free. You should leverage the test results in your recommendation. Also summarize any other main considerations that should weigh on the decision (what are the overall “pros” and “cons”?) — but make sure to reach a recommendation.
Feel free to review these materials and to ask me questions if needed. I hope this is largely self-explanatory, but if you do need to ask any questions, please send me an email. I know there are varying levels of depth with which you can respond to this case, but I am aiming for what can be accomplished in ~2-3 hours with the information I’ve provided and no additional research (or very minimal research). I’m very comfortable with you making assumptions for any data points you do not have in-hand as long as you make it clear which numbers are your assumptions.
Gallery case (mock data) (3) (1) Gallery case – sample images (3) (1) Gallery case description (3) (1)