1 00:00:06,450 --> 00:00:12,780 So that's, that's an overview to customer development, a bit of a taster and we're going to look at some theory now. 2 00:00:13,380 --> 00:00:17,280 So customer discovery, where does it sit in the customer developed model? 3 00:00:17,730 --> 00:00:21,480 And this is the customer development model of the theory. 4 00:00:21,790 --> 00:00:25,439 There are four parts. You can look this up, there are books about it. 5 00:00:25,440 --> 00:00:29,340 A great guy to look up who came up with this idea is Steve Blank. 6 00:00:31,830 --> 00:00:33,540 And we start off with customer discovery. 7 00:00:33,540 --> 00:00:40,620 So customer discovery is the phase where you're trying to work out if the people that you want to sell to exist, 8 00:00:41,430 --> 00:00:51,330 do they have the problems that you think they do? And can your solution help them solve those problems and see each of these phases as iterative? 9 00:00:51,510 --> 00:00:56,010 You shouldn't move on from one to the next until you've been successful. 10 00:00:56,520 --> 00:01:02,430 So until you've validated the fact that the people do exist and they have those problems and your solution can help them, 11 00:01:02,580 --> 00:01:08,490 you shouldn't move on to the next phase, which is customer validation, and that's can you build your product and get it to market. 12 00:01:08,670 --> 00:01:12,750 And once you get it there, will your customers actually buy it from you? 13 00:01:12,780 --> 00:01:17,810 Will they spend money on it? When she when she validated that that is all true. 14 00:01:18,200 --> 00:01:21,319 And then you move into customer creation and that's scaling your enterprise. 15 00:01:21,320 --> 00:01:23,360 So expanding it into the market. 16 00:01:23,360 --> 00:01:30,979 And then finally you stop that expansion and stop searching for your business model and start running the one that you've you've got in your hands. 17 00:01:30,980 --> 00:01:36,530 And that's company building and that's completely different. And today we're just going to look at the first one of these. 18 00:01:36,530 --> 00:01:40,429 That's customer discovery. It's where you'll start. It's really hard work. 19 00:01:40,430 --> 00:01:44,089 It's probably going to take up most of your time, and if you're successful there, 20 00:01:44,090 --> 00:01:52,630 I'm sure you'll buy the book and read about the the other phases from I actually urge you not spend too much time planning for those for 21 00:01:52,640 --> 00:01:59,000 those last couple of phases in particular have a good idea about how your company is going to run and how you're going to scale it, 22 00:01:59,420 --> 00:02:06,480 but don't do any preparation for it because until you've shown that you've got customers who will who will buy a product and it solves it, 23 00:02:06,500 --> 00:02:12,200 solve their problems, there's just no point going there. A couple of weeks ago, you heard about the business model canvas from Maria. 24 00:02:12,200 --> 00:02:17,600 So how does that fit in with this customer developed model? Well, you remember the canvas. 25 00:02:19,740 --> 00:02:25,830 Maria, explain that you should put Post-its on there explaining your ideas about how various aspects of your business are going to work. 26 00:02:25,830 --> 00:02:31,260 And it's a really neat way of describing and understanding the business that you're about to create. 27 00:02:32,250 --> 00:02:39,120 And if customer development is a way of systematically validating and adapting a business model, 28 00:02:39,720 --> 00:02:46,170 then the business model coming to us is at the heart of customer development. It's it's the tool that you used to run your customer development. 29 00:02:46,800 --> 00:02:51,330 So where does customer discovery live in that on the business model? 30 00:02:51,600 --> 00:03:00,420 Well, it's over on the right hand side and on the right hand side, you're talking in the see if I can do this value proposition section. 31 00:03:02,070 --> 00:03:07,950 You're talking about the needs that you're serving on your for your customers and how your solution is going to solve those needs. 32 00:03:08,250 --> 00:03:10,860 Over on the right hand side, you're describing who those people are. 33 00:03:11,930 --> 00:03:15,900 And then you're over here, you're talking about how are you going to get to them and. 34 00:03:16,910 --> 00:03:22,940 Well, in custom development, we're checking everything that you stated in this canvassing tree and changing ideas if it's not. 35 00:03:22,940 --> 00:03:26,810 So all of the stuff that you're writing down here is what we're going to start testing. 36 00:03:28,780 --> 00:03:35,620 And in customer discovery in particular, you're looking at and understanding whether or not your customers exist. 37 00:03:35,710 --> 00:03:39,220 Well, that's that's the customer segments over here. 38 00:03:40,000 --> 00:03:44,440 And then whether or not your solution is going to serve our needs, and that's in the value proposition here. 39 00:03:44,470 --> 00:03:50,110 So these two these two panels on the business model canvas are really where 40 00:03:50,110 --> 00:03:54,940 you'll start in in customer discovery and the fundamentals of the business. 41 00:03:56,320 --> 00:03:59,580 If you get these parts wrong, then there's just no point carrying on your debt in the water. 42 00:03:59,590 --> 00:04:04,540 So it makes complete sense if you think about it. To start with, who are my customers? 43 00:04:04,540 --> 00:04:08,290 What are their needs and how am I going to solve solving their needs? 44 00:04:10,240 --> 00:04:13,060 This may get slightly theoretical, but we'll see how it goes. 45 00:04:13,390 --> 00:04:20,380 And in his book, Steve Blank, this Start-Up Owner's manual, which you should probably all get, Steve Blank, 46 00:04:20,530 --> 00:04:26,260 suggests this is a way of thinking about how the business model canvas fits in with with customer development. 47 00:04:27,460 --> 00:04:31,900 So the canvas is at the bottom. It's your business model. It's your sketch of how your business is going to work. 48 00:04:32,560 --> 00:04:39,730 And on top of this, Steve suggests that you write a one or two page hypothesis for each of these panels. 49 00:04:40,180 --> 00:04:50,190 So one explaining in detail how your product is going to serve the needs and deliver value for the value propositions. 50 00:04:51,610 --> 00:04:57,909 And this can feel a bit onerous at the start, but I've definitely found it useful on the projects that I've been working on 51 00:04:57,910 --> 00:05:01,990 to go into more detail and just post-it notes that you get with the business, 52 00:05:02,590 --> 00:05:08,979 with the business model canvas, and you can pick the right parts of the business model canvas to expand on. 53 00:05:08,980 --> 00:05:12,310 Certainly those two that I spoke about earlier on, they're a good place to start. 54 00:05:12,790 --> 00:05:18,489 But you might you might find certainly at the beginning that some of them it's difficult to write more than the Post-its that you've got in there. 55 00:05:18,490 --> 00:05:22,180 So don't don't if it's really difficult, don't worry too much about it. 56 00:05:22,810 --> 00:05:26,860 Just get the essentials done. And that's the value propositions, the customer segments, the channels, the revenue. 57 00:05:29,140 --> 00:05:33,610 And then on top of that, you design tests to test your assumptions. 58 00:05:33,610 --> 00:05:40,390 Now, your hypothesis layer, you're basically stating all of your assumptions about how your business is going to work. 59 00:05:40,720 --> 00:05:46,080 So on top of that, you need to design tests that will then validate those assumptions and. 60 00:05:47,600 --> 00:05:51,020 It's a bit complicated. So I'm going to give you an example to show you how it would work. 61 00:05:53,570 --> 00:05:57,620 This is a project that I did with foils, the bookstore. 62 00:05:57,650 --> 00:06:04,560 And then the last year and what we were doing was creating a mobile search within the bookstore for, 63 00:06:04,940 --> 00:06:10,040 for people to search the inventory within that bookstore and find out if a book is in stock and if it is. 64 00:06:10,280 --> 00:06:16,160 Find a map to find that book. And it should sound like a neat idea is. 65 00:06:18,740 --> 00:06:25,490 This mobile shop search had a very simple value proposition, which was to the customers, search this shop on your mobile. 66 00:06:25,880 --> 00:06:31,940 Really, really simple. So let's think about that. The value value proposition hypothesis for this product. 67 00:06:33,460 --> 00:06:39,220 One of the and one way that makes it really easy to think about the value proposition is just to 68 00:06:39,220 --> 00:06:45,230 write a little story and write a bullet point story that explains how your product is going to work. 69 00:06:45,250 --> 00:06:52,230 So start by setting the scene. What's happening today? Explain the need of the of your customers here. 70 00:06:52,570 --> 00:07:00,970 Am I going to be using your product? Describe your solution to that need and then describe the results that you're expecting. 71 00:07:01,270 --> 00:07:06,429 And so here's that example that we've written out for FOIL. So today, many store customers need help finding a book. 72 00:07:06,430 --> 00:07:06,850 That's true. 73 00:07:06,850 --> 00:07:14,290 I've certainly been to a bookstore bookstore and gone to ask a bookseller if they can help me finding a particular book I'm looking for today. 74 00:07:14,290 --> 00:07:18,340 Some of those customers will leave without asking for help. For whatever reason, the queues might be too long. 75 00:07:19,720 --> 00:07:23,080 A mobile store search will appeal to some of those customers, 76 00:07:23,680 --> 00:07:30,220 and customers using the search will find that if a book is in stock and get a map to its location and then having found that book, 77 00:07:30,730 --> 00:07:34,630 customers are more likely to make a purchase. You can see this is really simple. 78 00:07:34,660 --> 00:07:40,360 It's just setting out in detail how that product is going to work and what value it's going to bring to its customers. 79 00:07:40,870 --> 00:07:46,510 And if you can't write something like this for your business, then your business is probably too complicated. 80 00:07:46,720 --> 00:07:52,990 If you can't convince it down into, you know, 5 to 10 bullet points, then you probably need to think about ways to simplify it. 81 00:07:54,130 --> 00:07:59,520 And it's also really easy to spot the assumptions when you've written it in this form and pick out the risk. 82 00:07:59,600 --> 00:08:03,850 So here's this one. A mobile store search will appeal to some customers. 83 00:08:03,850 --> 00:08:10,839 How how do we know? We haven't built it. We've got no idea whether or not anybody in going into that store is going to use our book search. 84 00:08:10,840 --> 00:08:15,310 So that start there when we're thinking about how to validate this business model. 85 00:08:16,000 --> 00:08:19,180 And you should also be able to spot clear need statements as well. 86 00:08:19,180 --> 00:08:25,229 So what the customers need. So having written our hypothesis and identified a risky assumption, 87 00:08:25,230 --> 00:08:32,790 it's time to design the and run the experiments to validate that assumption, to test whether or not that that assumption is true. 88 00:08:33,750 --> 00:08:40,590 So here's how we went about dealing with foils, and we put these signs up inside a store for one day. 89 00:08:41,280 --> 00:08:45,239 Over in Westgate, we took a Wi-Fi router, 90 00:08:45,240 --> 00:08:52,680 the sort that you might find in your and your house at home and put that in the store and called it and called the network the foils wi fi. 91 00:08:52,860 --> 00:08:57,720 And you can see that the sign says, log on to the foils wi fi to search the store on your phone. 92 00:08:58,260 --> 00:09:05,370 And when customers actually logged onto that, the wi fi and tempted to search, they saw a reply message saying It's time to come clean. 93 00:09:05,380 --> 00:09:09,480 We haven't actually built this yet. We're just testing it. Take this to the council and we'll give you 10% off. 94 00:09:09,960 --> 00:09:12,690 And we were counting the number of people who were coming into that store. 95 00:09:13,140 --> 00:09:15,900 And we were also came to the number of people who attempts to make the search. 96 00:09:16,140 --> 00:09:21,090 And we had expectations about how many of them we wanted to actually engage with our search. 97 00:09:21,390 --> 00:09:26,940 And that was a pass fail test. Either enough of them would do it, making the whole thing worthwhile, or they wouldn't. 98 00:09:26,940 --> 00:09:32,519 And the idea is flawed. And so we didn't need to build that search to validate that assumption. 99 00:09:32,520 --> 00:09:38,069 We just faked it. And your tests don't need to give you statistically significant results. 100 00:09:38,070 --> 00:09:46,470 This one did. Actually, we were able to work out with a confidence interval how valid the results were, 101 00:09:47,430 --> 00:09:50,010 but you're just looking for a strong signal that you're onto something. 102 00:09:50,310 --> 00:09:56,459 And some of the experiments that you run, the test that you run and will involve speaking to customers, 103 00:09:56,460 --> 00:10:01,290 some of them will involve faking a product like we've done. Some of them will involve building bits of your product. 104 00:10:01,710 --> 00:10:06,150 And what you're looking for is the simplest test that you can possibly run as quickly 105 00:10:06,150 --> 00:10:11,160 as possible for as little time and money as possible to test your ideas and move on. 106 00:10:14,800 --> 00:10:20,110 So here here are a few of the types of tests that you could think about running interviews with customers, 107 00:10:20,110 --> 00:10:23,919 and we'll talk a lot about that in a moment. Channel Activation Test. 108 00:10:23,920 --> 00:10:31,780 So ah, if I think I'm going to be able to market to people via direct mail, does that work? 109 00:10:32,890 --> 00:10:36,760 Sales channels are a great way of validating your business if you're actually going to sell something. 110 00:10:37,450 --> 00:10:39,879 Thanks. Products. We just spoken about hand-cranked products. 111 00:10:39,880 --> 00:10:50,860 So if you've got a product that involves lots of automated back office systems that are going to take a long time to build great expense, 112 00:10:51,100 --> 00:10:55,389 can you make it work with you doing the work of that automation, 113 00:10:55,390 --> 00:11:00,040 running around in the background, making making the business appear like it's working, 114 00:11:00,160 --> 00:11:08,270 but actually you haven't really built anything yet that's hand crunching. And then we've got to at the bottom, minimum viable products. 115 00:11:08,290 --> 00:11:13,960 These are basically simple versions of your products. We'll talk about those a lot later on. 116 00:11:15,510 --> 00:11:20,660 And so these are your methods, but really, you're only restricted by your imagination, you know? 117 00:11:20,700 --> 00:11:24,630 You know what you need to achieve. You need to do it quickly. You know, get creative. 118 00:11:25,510 --> 00:11:29,670 And when you get into the mindset of constantly, ask yourself, how can I test this? 119 00:11:30,150 --> 00:11:31,950 What am I solutions? How can I test them? 120 00:11:32,280 --> 00:11:39,930 You'll find it's enormous amounts of fun coming up with those ideas and extremely satisfying when you get a good idea for them. 121 00:11:43,180 --> 00:11:46,610 I'm and some some rules for designing those tests. 122 00:11:46,630 --> 00:11:49,630 Well, the very last thing you should do is build your product. 123 00:11:49,960 --> 00:11:55,120 I will admit that in some cases it will be unavoidable. But in most you shouldn't. 124 00:11:55,120 --> 00:11:56,020 You shouldn't have to. 125 00:11:56,950 --> 00:12:05,950 I find it really helpful in all of my work for clients and in the enterprises that I run to have this rule in the back of my head all the time. 126 00:12:06,220 --> 00:12:11,500 How do I not build this? I get my whole team to ask themselves, How do I not build this? 127 00:12:11,500 --> 00:12:17,500 How can I test this without building it and. The signals. 128 00:12:17,500 --> 00:12:22,170 So in your test, you're looking for a signal that the assumption is correct. 129 00:12:22,770 --> 00:12:26,879 And there could be anything that could be a trend that you notice in a number of customer interviews 130 00:12:26,880 --> 00:12:32,700 that could be a decent level of sales and they could be a statistically significant conversion rate. 131 00:12:33,090 --> 00:12:37,410 Don't get too up on the precision of the of the signal. 132 00:12:38,010 --> 00:12:39,810 It's not a randomised controlled trial. 133 00:12:40,050 --> 00:12:46,620 You just need to be objectively sure that you've seen the evidence to back up your assumption and then move on. 134 00:12:48,280 --> 00:12:52,540 If your test fails, you can still learn something along the way. 135 00:12:52,990 --> 00:12:57,670 And this is really important to think about when you're designing your tests. How can I get the most out of this test? 136 00:12:58,030 --> 00:13:01,660 What can I learn? Even if if a test shows me that my assumption is wrong? 137 00:13:02,020 --> 00:13:06,590 How do I maximise my learning from it? So think about that as you're designing the.