With the holiday season in full swing, retailers are leaning heavily on Buy Online, Pick Up In Store (BOPIS) services as an alternative to home delivery – and brands continue to prioritize omnichannel fulfillment as they head into 2020.
But implementing inventory transparency, store pickup, or store-to-online services within a seamless, personalized customer experience is more than a matter of flipping a switch. Jim Anderson, Kibo’s director of experience optimization, delves into why data is at the heart of a unified experience. This transcript has been lightly edited for length and clarity.
Omnichannel fulfillment and personalization are both top priorities for brands – but so far few brands are combining them. What are the challenges to successful implementation?
When we approach optimization, in addition to understanding goals and working to make the solution scalable, we want to look at it from a UX [user experience] perspective. So we say let’s slow this thing down, let’s put on our customer hat, let’s go through your Web site and then come up with some steps to getting into the solution. You have to have all the pieces in place – the UX on the Web site, the inventory management, the goal for the business, and does that goal tie in with me as a customer trying to shop the site?
For instance, I would go to a product page, and it would have me pick a store, and then the first message would be “Sorry, this is not in stock at your local store.” So right away I’m seeing a red flag that’s causing me to bounce, rather than saying, “Jim, this is available online, and let us check your store availability below – sorry, not available in Carlsbad, but we have it at another local store.” So just a simple change there can completely improve the experience and then all the engagement metrics. But that change involves a lot of key things: the UX requirements, and then the feed, and then layout and how it’s going to tie in and how we’re going to test the effectiveness of it.
What are the components of a quality feed? And how does that affect what retailers can offer to shoppers?
You’ve got product availability, and if it changes multiple times throughout the day, then you need a delta feed [reflecting changes since the last update] to make sure we’re giving the most accurate picture for store pickup. Going back to the out-of-stock example, merchants say, we update inventory once a day, a lot of times we’re still recommending stuff at the end of the day that we don’t have but are showing up on the site. Okay, let’s use that delta feed.
Also, let’s make sure that your feed has the right attributes, so that if a shopper is looking for a 17.5”x34” pink shirt, if we don’t have that shirt, we can pull metadata or attributes from what the shopper is looking at to say, “Sorry, we’re out of stock, but click here for more like this.” If I’m the shopper and I’m in a hurry, and I click the button and I see three or four other shirts that are a similar brand or have similar attributes, that’s going to help me complete the journey. It sounds pretty simple, but if your feed doesn’t have the data, then it’s going to be hard to recommend like items. So much comes back to the feed.
What are some feed attributes that might not be obvious, but that merchants should consider when starting implementations?
One is shipping rules. Sometimes you’ve got, say, an automotive client, and they’ve got items that ship in a small box, items that ship in a big box, and tires. If shoppers hit a certain threshold with small-box items and then all of a sudden the retailer recommends this really inexpensive big-box item, they’ve got free shipping because they’ve hit the threshold, and we’ve presented them something that goes together but is now costing the retailer money.
The seller would need to flag and let us know, what are the situations where this could happen? When shoppers hit this threshold with small-box items, don’t offer big-box items or tires. Give us “small”, “big”, and “tires” in the feed, and we can write conditions that allow the recommendations to work and make it more business-centric.
Another is margin optimization. If you have two like products, you want to optimize based on margin. But there’s a law of diminishing returns; if you just use margin, then you’re taking out the affinity scoring of, this product works with that product, and what products should be at the top of the list based on the shopper’s profile.
If you use margin at 100 percent, the shopper says, “These aren’t as applicable to me, great I’m glad they’re higher margin for you, but they don’t solve my need, so I’m not going to engage.” But let’s test it and see what the dropoff is. If the dropoff is only 5 or 10 percent, the merchant may say, “I’ll take that drop if I’m going to get to drive 40 percent more margin.”
You touched on an important point with measuring effectiveness. When shoppers jump from browsing online to the store experience, how can retailers capture that?
If we can’t tie online and offline together, then let’s at least see their last online touchpoint, that’s different from the old flow prior to the implementation, and let’s look at the impact there. Maybe now we’re down 200 conversions per day on our page. But 180 of those are clicking on this new option, and we find that when people go into the store, they spend twice as much per order than those other 20 orders we just lost, so the net gain is significant. It’s about key events. We’re talking about doing something you haven’t done before, so how are we going to judge a click and success?