Let's talk about a few ways you can do that on incremental features. Okay, here's a mockup of what the Enable Quiz might look like. One really interesting test, and it's just one tool, is to do a fake feature test. So let's say this is their normal menu for adding and administering quizzes. And they add this little tab here, Skills Audit, and they see if anybody clicks on it. This is on a SaaS product where they control the software, and they have a lot of users. They only roll this out to a subset of them, and they just see who clicks on it. And if they click on it, they come to a tab that says, Coming Soon, or sign up for updates here if you want them. And this is a good way to gauge interest under these particular circumstances. So if a lot of people click on it, what can you conclude? Well, you can't conclude that they're going to love it and use it. But you can conclude that they would take note of it and check it out, and that's good. That's a positive initial signal that there's some interest. And the thing that I wouldn't do is overconclude about this. You also don't need a huge amount of statistical significance here. You just want a notion of whether people will click on it. So this fake feature task is a good thing to do. Don't over-conclude and don't run it too much or too long. You just want to get a relatively small sample to see if the people are going to click through on it and take note. Another thing to do, a great tool for new features, is to concierge them. So just like we talked about with a whole new product, we take this incremental thing we're thinking about adding to an existing product. And we do kind of a hand-carried version of it. I know it seems like a lot of work. But if you think about it, all software can do, at the moment anyway, is automate and standardize things that we already know how to do. There's machine intelligence and software's learning how to be its own creator to a small degree and that's certainly coming. But right now mainly we standardize,and automate, and structure things that we already know how to do. So if you think about it, the idea that we don't really know how to do this and we don't exactly know how customers do it. But we're going to build software for it, it's pretty crazy when you think about it. So I think it makes a lot of sense to hand create this experience. It's not going to feel like real work. You're going to wonder if you should do it, you should do it. So another great tool for the Enable Quiz team is to concierge the Skills Audit, so find a customer that wants to do this. How would you do that? Well, maybe you ask them a series of non-leading questions at increasing levels of specificity. So for example, hey, tell me about what's on your A-list management-wise to the functional manager that takes care of a bunch of engineers. And what have you done in the last few months that really helped you do a better job as a manager? What would you do if time and money weren't an issue? Now, if they start saying things like, I get everybody on an e-learning platform. And kind of help them tailor their own programs so they could advance their skillsets and move their career arc the way that they want. Then that's a good signal that they might care about what we have. And if they don't say anything about it, then we don't want to concierge with these people. We might not want to concierge with anybody. If this problem doesn't exist, we don't want to pursue it. So this is a good thing to find out early, because that allows us to pivot and focus somewhere else on something that does matter to them. And so, a good way to find a customer to participate or customers is to ask them a series of non-leading questions, something you should do anyway. If you remember we move from persona to problem scenario, who is this person? What do they care about? And then Enable Quiz maybe just does this whole thing by hand. So we find out what kind of jobs they want these developers to be able to do, which ones are important. They design assessments to test for that with their quizzes, they do those. And maybe this is or isn't on the actual platform. That's less important because we're not on a scaling mission with this. We're not trying to build the 1.0, the Skills Audit. We're on a learning mission to learn if and what this thing would be to be valuable to the customer. They would take this forward to a professional ed program. Yes, they don't sell to professional ed. But if that's not the outcome that this thing can drive to, then we think that they're probably not going to care. Hey, it's great, they have these skills. But what it has to be, we think probably, our hypothesis is that it has to be actionable. So we find the professional ed program even though we know that's not in our area. And we help them run a pilot to go out and upscale the employees. And so, this seems like a ton of work. But again I think personally it's crazy to try and standardize and automate something with software that you don't understand really well. So concierge testing is one of the most powerful vehicles you can use to increment your features on the margin. It may not be a right fit for your particular product. Maybe the fake feature test is, maybe something else. But this is a great way to apply the methods that you've been learning about in the course on the margin as you develop new features.