Wrapping Up

[00:00:00]
>> So that's basically it for testing an existing site is either AB testing or prototyping again, basically, post launch testing and iteration are probably the most valuable thing you can do. In everything I've talked about today that kind of ongoing makes such a difference. Having that kind of backlog of friction points that you add to regularly is really good.

[00:00:23]
So rather than just having them sit in analytics, you might actually wanna create some kind of backlog with them all in. For more minor fixes, just use a bit of A-B testing. For larger fixes, consider prototyping and testing first before you commit to something. Very last thing I wanna say, cuz I haven't said it for a while, is I've shared a lot of stuff with you, but just do what you can.

[00:00:48]
Something's better than nothing. And if all you do out of every in this entire workshop is run a semantic differential survey or run a first click test. And that's all you do. Then that's great. That's a step forward. And it's beginning to introduce those kinds of things. Personally, I find this kind of lightweight usability testing really valuable.

[00:01:15]
And it saves me time ultimately and it saves me so much hustle with clients and stakeholders that I do most of it out of my own pocket. I mean, it's got better over the years and now I've got better at persuading people to actually part with money for it.

[00:01:33]
But for a long time I would just do it because it made my life easier. And hopefully it makes your life easier too. So what we've covered just to summarise what we've covered today we looked at a set of principles that kind of underpin lean, lean UX. We've looked at doing upfront user research and how it can be used to create an information architecture, how we can do some basic design testing around design.

[00:01:59]
Comps and in many ways, I think that's the most valuable part of everything we've covered. We talked about opportunities to do usability testing with prototyping, and then the real powerhouse that's testing existing services over the long term. But as I say on here, and I 100% mean it, I really enjoy sharing this kind of stuff with people because I know how painful it can be to make testing happen and how painful the design process can be.

[00:02:27]
And so if there's anything I can do to help, by all means, email me anytime my email addresses in the slides pull up by google.com and I'll happily answer questions in six months or a year's time, it makes no difference to me. But are there any closing questions before we wrap up today?

[00:02:47]
Yes?
>> So early on in the principle section, you mentioned how to translate the initial effort into actual monetary value for the business.
>> Yes.
>> Time or actual money saved.
>> Yes.
>> How do you go about doing that?
>> That's a good question. It's a lot of guesswork, right?

[00:03:09]
But it is possible more often than you think. So let's say, I know, I'll give you a real, it's not a real example because I was making it all up. I was in Pizza Express one day, which is like a chain restaurant in the UK. And the waitress came over to me and she'd taken our order and it was taking her forever, put into an iPad, put this order in.

[00:03:35]
And my wife could see, she couldn't now mind read me. We'd been married so long and she went, no. And I was mentally counting how long it was taking her to place the order. And when she went away, I said, right. Let's say she knew, I knew she knew, and it was one of those weird ones.

[00:03:55]
But I said, what if we could shave 10 seconds off of it for every order, right? And then I started making guesses, right? And it's all guesses, I said, right? Let's say the average number of people at a table is four, right? And let's say they do X number of servings per day in this restaurant.

[00:04:21]
Then I looked up on Wikipedia [INAUDIBLE] life. I looked up how many restaurants they have, right? How many days they're open a year. And so I could calculate how many seconds that added up to over a calendar year. Then I worked on the basis of how much a person earned, yeah, I put minimum wage for a serving staff, and then I could calculate how much I could save the company by taking off 10 seconds off of each order.

[00:04:55]
So all of that was ghetto except for how many restaurants there were, in which I looked up, all the rest of it was guesswork, right? But that was fine. And if they turned into a client, they would have corrected me, and I would have adjusted those numbers, and we would have seen whether the project was still worth it.

[00:05:14]
But doing that kind of thing is incredibly valuable, because it Attaches a monetary value to what you're doing. And everything we do will have some kind of benefit like that. That was a cost saving benefit. Sometimes it might be a customer retention benefit. So for example, let's take Frontend Masters, right?

[00:05:37]
So they've got customer acquisition that they've got to meet. And so stuff on their frontend website would be if you made improvements on the front end website, that would help customer acquisition. So we could say, well, if we could increase the conversion rate of people visiting the website by 1% over a year, we could calculate how much that would be worth to the organization, right?

[00:06:00]
It would be guesses, but better than nothing, or we could focus on the back-end app where people log in and use the courses. Now we're talking about customer retention. So we could calculate, well, if we could keep people for a year longer than we were previously, how much would that equate to into?

[00:06:18]
In terms of extra profit. So it's making shit up but but it's educated making stuff up. And you're not claiming these figures are real but it's starting a conversation about what the real figures might be. And it's starting to put tangible value on what it Is that we're doing.

[00:06:39]
Does that make sense? Yeah, go for it.
>> I sometimes see websites and apps that have a regular feature to provide feedback.
>> Yes.
>> What's your experience of the value of that?
>> Sure, so there are two types. One of my favorites, I run a lot of one-off surveys.

[00:07:06]
That's different to what you're talking about. But I do find those very valuable. So one of my favorites to run is a one question survey, cuz like I said, I do a lot of conversion rate optimization work. So I put one question It only triggers as a popup on exit intent.

[00:07:22]
So as people are about to leave the website, and it says if you decided not to act today, it would be great to know why, right? And here's a load of different options that one question can be transformative. So there's that kind of thing. What you're talking about his were at the bottom of the pages.

[00:07:38]
So did you dind this page helpful, right? That kind of thing. Is that what you're getting at?
>> Or even there might just be something like there'll be a smiley face or a frowny face and just you're.
>> Yeah, that kind of, yeah.
>> It's really open-ended.
>> Yeah, yeah

[00:07:50]
>> It's not soliciting [INAUDIBLE].
>> So those kinds of things, they actually have their worth, but, probably not so much from a user interface point of view but more from a content point of view. I use them a lot when I'm working to improve the content on a website because one of the big problems that happens in a lot of organizations is that content production is spread across the organization.

[00:08:21]
So different departments, different people, they own different pages on the website and they've got to update it. And it's just an extra part of their job and they don't care about it. And so the content is rubbish and there's no feedback on it or whatever. But I often create dashboards that will return the results of those, was this page useful, smiley face, happy face kind of thing to the owners of those pages so that they can see how their page is performing versus other people.

[00:08:56]
Now, suddenly they start caring when their pages are scoring really bad, but their colleague in the next department is scoring really well. They don't like that, and that makes them improve their content. So it can be a good internal motivator for stakeholders to actually maintain their content. That's where I find it most useful.

[00:09:17]
So yeah, it has its place but maybe less so from a usability point of view cuz it doesn't give us a lot of information to go on.
>> Yeah, I have to say we we actually implemented virtual assistant in the mobile application. And there were two types of servers that we introduced.

[00:09:37]
If you close the survey, I'm sorry, if you close virtual assistant, you'll be prompted for a survey to say, how do you like virtual assistant? So it's feature specific in the mobile application versus the whole mobile application.
>> Yeah.
>> And then we had another survey where if you didn't wanna talk to virtual assistant or a bot, you wanna talk to a real person, you can do that.

[00:09:57]
You say, I wanna talk to a live agent, which is really a banker. And then the banker answers your questions. And by the time you're done with that session, inside live agent, you'll be prompted with a survey that says how I was working with a banker in live agent.

[00:10:11]
So there's multiple categories of surveys that we use to collect whether or not Is this working for us? If it's not, what can we do to iterate and make it better? But we initially planned for that upfront.
>> I think surveys can be incredibly valuable if it goes back to what I was saying earlier, you need to actually do something with the data.

[00:10:35]
You see so many people take surveys and everyone sits around afterwards to go, ooh, that's interesting. And then it gets put in a drawer and forgotten. And all it's done is annoy users, [LAUGH] right? Cuz that's the danger, you've got a balance the whole time. But in those kinds of examples, in especially the second example where you've dealt with a person, that is feedback is valuable to that person.

[00:10:58]
It helps them with what they're doing. And so in those kinds of cases, it's really good. But you got to be careful with surveys that they're not getting the users aren't being bombarded with them again and again and again. Every time they go on the app, the survey's there again.

[00:11:13]
That kind of thing, you've got to be careful that they're only seeing it once, etc. So yeah.
>> There was a new feature that was made available on the New York Times app website, or New York Times app that I use, and it would read the article in a higher quality voice than the built-in

[00:11:31]
>> Yeah
>> IOS text-to-speech, and I had some feedback for them. And I was able kind of tap something to say what is this? And it took me a few taps to get there.
>> Yeah.
>> And I had to type in your message and I did get a good response but I am curious like how many people go through that trouble and and is it I can tell by their response that they were interested?

[00:11:54]
>> Yeah.
>> So it must have been some valuable but I'm curious how valuable is It's like, how common that is.
>> It's more common than you think. The only thing you've gotta be a little bit of aware of with things like that is that generally speaking the people that respond to surveys, either absolutely despise the thing or absolutely love it.

[00:12:17]
If you're ambivalent about something, you don't bother filling in a form, right? It's only if you really love it or really hate it. But presuming you understand that basic bias, it's worthwhile data. You get a good number of responses. In fact, I've got a problem at the moment.

[00:12:34]
I'm working with a charity where we're optimizing engagement funnel where we're sending emails out to people that agreed to fundraise, and we've encouraged them to respond, right? And actually they've overwhelmed us. I completely caught me off guard because we wanna respond to them and we can't keep up and we were trying to get people in to respond to all of these these messages.

[00:13:04]
So sometimes you get the opposite problem. [LAUGH] That's quite unusual minor. I've had that for a while. I should call it day. Thank you, everybody for for joining. It's been really nice to have actual people in here. Last time I did this we were kind of the end of COVID and then it was a bit kind of me on my own some.

[00:13:24]
Thank you for those people that joined online as well and that have watched the course so hopefully it's been helpful.
>> [APPLAUSE]