How Anticipating Unseen Customer Needs Transformed Expedia's User Experience

OK. Yeah. So I do have an example of this, and um it's gonna be a slight uh variation of what you've asked for. It's um definitely a case where I anticipated a customer need or in this case, a customer problem which, which not too many people saw, and I, uh, handled handled it appropriately. Is that, is that OK? That's fine. OK, so, um, I'm gonna talk about, um, this experience that I had at Expedia. Um, to set the context, uh, Expedia had hired me to head the, uh, user identification team. So we own the front end and the back end of all the authentication experiences, and it was our remit to increase the, uh, number of identified users on the site. And one of the initiatives that I had zeroed in on was implementing uh Google AutoSignin. So you have seen this on multiple sites. You land on the site and then it, uh, there's a little window that comes up on the right-hand side of the site that says continue as aju at gmail.com or whatever you click on it and you are an account is created automatically. Yeah, so this, yeah, so this seemed like, uh, from user studies we knew that customers looked at account creation as a burden with no immediate gratification. So, uh, it, it's a painful and create account. So we thought this would be an extremely good, uh, feature to get more customers to create accounts. So we were very optimistic about it, um, and, and before implementing it, I did my due diligence and thought about what would happen if we, if we implemented it. Everything seemed to point in a positive direction. And on giving it some thought, what I realized was, supposing you're a customer who say, you had Brandon at yahoo.com, um, you've created an, uh, you have an account at Expedia with Brandon@yahoo.com, and then you bought an airline ticket with us. And then after that, you visited us and you saw this, you visited us after we implemented the Google Signup. You see this little easy to use button there, and then you click it, and then now you have a different account with us, Brandon at gmail.com. Now, imagine you're at the airport and you're trying to pull up your item, which is associated with the with your Yahoo address. Right, so you're trying to find, you, you'll be logged into Gmail automatically, your Gmail account automatically at Expedia, and you won't find your item. It is a terrible experience, right? So, at first blush, this seemed like something that'd be great for all customers, but I'm digging deep, I found that there'll be some cases where we call it the account bifurcation problem. So, We found that uh um uh customers are likely to face this account bifurcation problem if um we implemented this and so what I did um to to chart out next steps where I thought this is significant enough it's a one-way door but it's significant enough for us to run an initial test to get some signal as to how we're doing. Right, so what I did was I worked with the customer service team and I worked with them on a script for any customers who may call saying that, hey, I'm not finding my IPI in my account, right? So I made them aware that this is one thing that could have gone wrong that the customer service rep should help them with. After that was done, we ran a test, um, uh, with about 5% treatment, uh, where we showed customers this particular, um, um, you know, the Google autoign an experience, how to sign up sign an experience. And then we monitored to see if any uh uh calls, the volume of calls that came in with this kind of account confusion uh case. And once we found that the volume was low, then uh we um went ahead with the rollout. And uh we thought through the case. I thought through the case when if we found that there were a lot of account, um, what do you call bifurcation cases within that 5%. I thought through the next steps. So in the, in those cases, we would have, uh, we had a plan to move people to, uh, uh, we would have simply removed the, um, auto sign in, um, experience, and then, uh, those customers would have had to sign in, uh, regularly, uh, uh, through the regular means, and, uh, the auto auto sign up functionality will be removed, right? So that was the, the, uh, part that we had in case, uh, this didn't work out. And we had um like a a guardrail metric, a certain number of calls if they were reached, then we would scrap it, otherwise we would, uh, go ahead with it. And ultimately we found that it did not impact a significant, we did not cross the guard rail. A number of customers were impacted and we implemented, um, Google Auto Sign in which drove in account creations, uh, in the pages where it was implemented and drove in by about 50%, uh, account

uh, 50.

cases? Yeah, I should have given you more detail about that. So this was not a fixable thing. So supposing in the example that I gave you, you have an account, Brandon at yahoo.com, and then you automatically, we automatically create Brandon at gmail.com, a separate account at Expedia for you. Now there's no way for us to know what your Yahoo address is. So for us to do something like merge accounts and things like that, uh, there's no way for us to detect that. And, um, that's the reason why this is not a fixable problem. Um, does that answer your question?

So, um, at Expedia, I don't know, uh, I don't think like it's been about 1.5 years since I worked there, and as far as I know even today you cannot, if you have 2 different accounts. Uh, there's not a way for you to log into the same account. Um, some other sites may have, um, means of merging the accounts. Uh, that is if you use the same underlying email ID. So you can have, um, your email ID with your Facebook account can be your same Gmail ID, right? So, a lot of people do that. So in that case, we'll be able to detect it on, on, uh, behind the scenes and, uh, do a merger account for you. But in the case that I mentioned where you, you're using a Yahoo, uh, email address, for example, And your Gmail, Google uh address is bound to be a Gmail one. in that case, it's, it's, there's no way for us to detect and merge. OK, and there's no way for us to merge without involving the customer automatically merging is, uh, we wouldn't be able to do that. We have to do it with the customers.

Yeah, so, from, um, uh, customer, uh, research studies, what I resoundingly heard was that customers just don't want to create an account and sign up, um, in the context of, uh, Expedia especially. So we created, um, several sign sign up experiences, several value props, like, you know, get, get loyalty points, convenience, and so on. And what we found was, you know, customers just don't wanna do it, right? So the takeaway from, uh, for me, the takeaway was my job is not to build the most fancy sign in set up module. If I can get out of the customer's way as much as possible, then that's what's gonna help the customer. That's what's gonna help me be successful. So that's why, uh, automatic Google account creation and sign in was such a good, such a, such a great option for us because of customer convenience. We had other means to do it automatically as well, but this was one of the tools in our toolbox.

You as a customer, you have to go to the top of the site, top right corner and navigate there you click on sign up and then you have to deliberate deliberately click on create account, right? So we found that very few customers actually bother to do that. There's, um, um, if, if I remember right, it was some 3 or 4% of all unique visitors bothered to do, do that. They just are not motivated to do it. Um, and another, uh, so when we implemented the automatic, uh, Google account creation. It, uh, initially actually we anticipated even more impact, but the impact was limited because a lot, a lot of customers have disabled, uh, JavaScript on their, uh, browsers, uh, quite a large number, uh, more, more than I anticipated have actually disabled, uh, JavaScript on their browser. So for those customers, you don't have, um, this particular module coming up and, um, of the remaining quite a few, like I, I forget the exact percentage who. Uh, created an account, but, um, we got a significant number of customers creating accounts and that is what led to the 50% increase, increase in, uh, completion, to be honest, we expected to be much, much more than 50%.

right, yeah, so that's a very good question. So, um, we, we have, um, I had converted, I had attached a dollar amount to account, the actions of our account creation and, uh, sign in. So this would be um um in the region of, OK, I, I don't get those numbers, but um for this experiment as well we went through, we went through the exercise of associating a dollar dollar amount. And we look at it over a year because if you um look at the purchase frequency of Expedia or a company like Expedia, it's like there's people buy typically something on the site uh 1.something times a year. So really to understand the impact of something like this, you'd have to do a cohort analysis over a year. To see what you're having. And we did in fact do that and we saw an incremental lift in terms of the dollar amount that you can associate with uh account creation. So customers who, um, visited the page on the same day, who did not create an account and customers who created an account because of experience, if you look at the difference between those two cohorts over a year, we did see an incremental lift. But it was not as much as, as you rightly pointed out, it was not as much as somebody who navigated to an accomplation page on their own accord and created an account. So there's definitely that, that difference. OK.