Businesses usually want growth. One revelation of JTBD theory is you cannot just think about product, you have to think about your entire offering.
When buyers are shopping, they’re considering messaging, branding, features, price – all of it together. So when you’re designing your next product offering, you have to test all of them together. Alan will show how you can approach this complex challenge by breaking the problem down and designing meaningful experiments to test your hypotheses and assumptions across the organisation so that your growth can be more predictable. Using real examples including Apple and Lego to demonstrate how to get better results, Alan helps you think about how to optimise for current customers and invent for the future.
Learn how great SaaS & software companies are run
We produce exceptional conferences & content that will help you build better products & companies.
Join our friendly list for event updates, ideas & inspiration.
Today, I want to talk about my journey and experience with using jobs theory to making growth more predictable. I threw the more on there because we all want growth to be predictable. But the reality is, we can only make it more more predictable, we can’t have the perfection, but we can still strive towards that. So this is my my journey here about using jobs theory for for more predictable growth.
So if you’ve been following my work for a while, you’ll notice that for the last three years, especially I haven’t been very vocal about jobs and jobs theory, I’ve been writing about it. People have been asking me about another book, so and so forth. And there’s a reason for it. That’s because about three years ago, I realised for me, there was a giant hole in jobs theory; something really huge was missing. Long story short, what was that? It was prediction. I was using jobs theory to understand how and why consumers are buying products today. So what were their needs today? And how are they attaching those needs to current things in the product? What was the competition, so and so forth? It’s helpful, for sure. It’s great for optimising things, but for creating growth, especially more predictable growth, it has limited use.
And to kind of give an idea, think about weather theory, what’s the value of weather theory? Well, if all weather theory did was tell you, if you look outside, right now, it’s a hurricane. Well, that’s not very helpful, I can look outside and see that’s a hurricane. Weather theory is helpful is when it predicts and it tells me, for example, if you live in Miami, you’ll have a 30% chance of every year of being exposed to a hurricane, or there’s a hurricane right now, and it’s gonna hit you in two weeks. Being able to predict the future enables us to make decisions today. So we can realise a future that is most beneficial to us. Same thing for plate tectonic theory. It’s great to know, it’s fun, and then an elementary school to teach those kids that there was once Pangea, you know, all those continents were together, it’s fun to teach them how the Himalayas were made. But how useful is that?
What’s more useful is predicting where an earthquake will be, when it will happen, and what the intensity will be, again, so we can plan for the future. And, of course, the crown jewel of science and prediction, which is physics. You know, we didn’t A B test our way the moon, we didn’t MVP our way the moon, you had to get it right the first time. You know, so and that was done through great many predictions predicting where’s the earth going to be in relation to the moon? How much fuel we’re going to need? So we escaped the Earth’s gravitational pull? How much oxygen will we need? You know, at the time, they weren’t even sure 100% if we’re if an astronauts bouncing around in jumps, will he fly from outer space? So they had to predict how much weight do we have to have in the spacesuit so on so forth, you get the idea. So this is why prediction is so necessary.
And so when we ask this question of well, what, what job? Or what other job should jobs theory be doing for us? I will say, it should enable us to understand and predict what will cause consumers to hire products from us. And if they won’t, why? So what we’re gonna kind of unpack that statement. It’s like answering questions such as which products should we bring to market? And how should they be designed? And what should our go to market strategy be? How should we position and message these these new products? What will work? What won’t? Will it change in price packaging or features attract more customers or not? And or even something as simple as outbound? Who should we contact in our outbound leads? When should we contact? What should we say to them? Those are all questions of prediction.
So that’s for me, I’ve last three years, my team and I have just been intensely focused on how can we advance enrich job theory so we can predict hiring. So I’ll give an example here of a real life project. So about two years ago, we work this company Dream and real quick introduction. They make incredible technology that will analyse how you sleep, and it can tell you and endless things right? And in fact, there were so many things they could do with the technology, they weren’t really sure. It’s like there’s a million things that we could do and tell users about how they sleep when they should get up when they you know, what kind of I mean, even diet to an extent. So, you know, there’s there’s that one problem of where do we take the product next to advance it. But the other really more pressing growth problem was, they were really good at capturing innovators and early adopters, I don’t know if you see my mouse or not, but it’s hovering over innovators and early adopters, but they were unable to diffuse to the market, they just could not crack into early majority, and definitely not late majority and laggards.
So of course, this is this is prime jobs theory, right? What will, what does Dream have to put into the market that will cause consumers to buy a product from them? And if they won’t buy it, why or why not? So we work with them to come up with two growth hypotheses. One was one, they were already we’re pretty sure of it’s they thought, Okay, we we believe that people who are have an inkling idea that they have some sort of sleep disorder, like an apnea, or I can’t call the others but there’s, there’s a whole laundry list of brain functions or things that your body does that can that can disrupt your sleep. That was accurate, but they thought, okay, you know, even if we, you know, they do the TAM analysis and so, so, so forth. They’re like look, we need more growth, we got we got to sell more of these things. So, they have this other idea of okay, well, maybe these these life, we call them lifestyle optimizers. And these people who are interested in you know, people like buying Aura ringing or you know, the Fitbit, people of that nature, that okay, what can we put into the market, so that they will want to buy a product from us the Dream product from us.
So, long story short, the first group of ICPs of target consumers slam dunk, no problem. The biggest reason was, you know, they, you’ll see later on talk more about it, but it just kind of hit all the notes of of jobs theory, that they can recognise it, they can understand it, they can attach a job to done the product. That made sense to them. But also really good, it was really good value for money. Because you’ll see later on that the price, it’s like, oh, that price, that’s totally worth it, because I spend twice that much each month on therapies, and so and so forth. But this group, nothing. We tried so many permutations of positioning messaging, pricing features, you name it, we just could not generate willingness to hire amongst these ICPs. I’ll talk a bit more about that why later on. But the point I want to come to is that this is really, I feel like the sweet spot of jobs theory making these predictions in advance that no matter what you do dream, you will never attract these types of customers, it doesn’t matter what what you price it or the features, it will never happen. So with that decision, they executed a major pivot. And so now they’re focusing on they’ve dropped direct to consumer. And now they’ve pivoted to the model of a b2b play. So to me, that is, again, the sweet spot of the application of jobs theory, being able to predict why a group of people will or will not buy a product or some permutation of your product offering this even before MVP, for example, even before that.
So we’re going to talk about today is, well, how do we construct those predictions? And how can we do lots of predictions? So of course, my answer for today is well, jobs theory and a method I’ve developed called simulated selection. So today, I’m going to do just just enough theory, you know, because, you know, I could go on and on about it. Actually, I am developing a video series right now. It’ll be I’m not sure if it’d be on Vimeo or YouTube. But it’ll be it’ll be free. That’d be out there. And right now, we’re at about 25 videos, just about jobs theory, kind of really going into all the weeds of the needs, lifecycle goal, homeostasis, discrepancies, catalysts, categories of catalysts, constructing models, dependence pathways, whatever, just we really go deep in all that to help you get really make you become your best at make using Jobs theory to make predictions, but that’s beyond the purview of today. I want to focus on just these two topics, which is the tip of the iceberg and arguably the practice of of making up prediction, which is willingness to hire and the hiring process.
What is the hiring process, or at least my interpretation of the hiring process? So, the official definition I’m going with is to process the application of various heuristics and decision making processes. So the shopper can determine their willingness to hire a product. What does that mean? I’ll just kind of cut to the chase here, basically means that we, in our research and studying consumer behaviour, and running experiments, and so on and so forth, we’ve realised that there’s about five decisions. I mean, you know, we had to come heuristics, but I want to avoid the jargon for right now, there’s four variables that shoppers use, to construct their willingness to hire. So the first one is recall, relevance recognition.
Second one is job to be done. Trust, use and value for money. So I’m going to unpack each of those right now. Or, we also like to, in our minds, think of it as a formula. So willingness to hire your product is some function of relevance, recognition, jobs done, use trust and value for money. So going through each of these real quick, relevance recognition, what do I mean by that? What I mean is that this is how the shopper determines if the product is relevant to them. And also, if it’s new, actually, it probably should be reversed if a product is new to them. And if it is new to them, is it relevant to them? And I’ll give you an example here is, if you go look at the new iPhone, I’ll kind of zip through these kind of three, three images here, you’ll find on the iPhone 13. Homepage. It’s very interesting to me how Apple is introducing the product by showing the back of the product. But it’s kind of interesting, right? It’s like showing the like, actually, I think Audi does that sometimes in the past, but they show the back of the car doesn’t show the back of the product. What’s Apple doing that? Why are they doing that? Well, the trend trigger recognition, and relevancy. What you’re saying here is, this is a new camera system. Pay attention. This is a new camera. That’s what they’re saying. And all of these right, and probably alluding to new screens, there’s a there’s a new screen, and a brand new camera system. And also, look, we have different variations of this camera system to you. So this is how they’re trying to trigger the shopper to continue the hiring process. It’s think of it as like the hook in their mouth that says, oh, that’s new. And wait a minute, I care about pictures. So you’re not really I want to know what it does. But already I can tell this product is new. And it’s relevant to me just by simple picture. So that’s when we think about the recognition, and Robin’s heuristic. And we’ve actually mapped out the pathway of this, which helps us debug it.
So in this case, for example, the consumer gets prompt as a prompt, or catalyst, where the consumer is exposed to a product that can happen any number of ways they see walking down the street, they see someone using a product, they see an ad for it, the friend uses it or the friend shows it to them with some sort of introduction to the product. So the first thing they see, firstly, it triggers something called a comparator. If you’re curious about that, just type in like cybernetics comparator, it’s basically just a function that humans or any decision making process does to compare. So the consumers thinking, Wait a minute, have I seen this with a shopper? Have I seen this before? If they feel like they have seen it before, we’re going to pull from memory, their previously formed conclusion of it, and then move on. But if they determine Wait a minute, I haven’t seen this before, then another comparator fires, and they think, Wait a minute, there’s anything about this product seem new to me. If they determined that nothing about is new, then they’re actually going to store this new conclusion about this product in their memory, and then move on. But if they do determine that it is new, and relevant, then they’re going to continue on the hiring process. This is really important to you, I heard someone talking about building their web page. You know, this is really important to understand. So when you’re testing with users or thinking about your new webpage, for example, you gotta make sure that you’re you’re passing this recognition and relevance heuristic. And this can be for ads too or even the look of your product.
Because, you know, again, the shopper, you know, I’m sure there’s some number 1.5 seconds or something like that. Well, that’s what this happens, right? If someone wants to continue looking your product or not. This is also relevant, and that a lot of times shoppers will misinterpret your product. They’ll think oh, I’ve seen that before. I know what that is already. Know, I don’t want it, but it could be a misinterpretation. So it helps when you have to kind of work with or test and experiment with shoppers. And debug this process. If shoppers are determining that it is not new, but when it really is or that it’s not relevant, but it really is, what can you do to fix that?
So next, next heuristic is usable. The belief the shopper that they can actually use it, or others will use it. I don’t spend much time on this. This is basically just as it sounds. Except that will add that it’s not just like usability, which we often think about as product people or designers like UX. It’s literally like, will I use it? An example is we did a study with peloton, shoppers, consumers about two years ago, right started the pandemic. And when we talk with them, when we said well, you know why? You were familiar with peloton in the past? Why didn’t you buy one in the past? You know, because most of the conclusions were the same? Yeah, you know, I like what it does for me, and I trust the brand, so on and so forth. But you know, why now, and it was always well, two things, there was a value for money calculation, which I can talk about later on. But they thought, well, actually, I will use it. Now. You know, before I anticipated about the peloton bike, I get my gym membership, or sometimes our workout my friends, so I probably wouldn’t use it like twice a week. So, you know, that’s not really worth the value for money. If I’m only using it twice a week, Yeah, feels on my needs, and so on so forth. But it’s not worth it for twice a week.
Another way of thinking about usable is some people said this, I couldn’t even get into my house. I wasn’t even sure if it’d fit in my house. So we think about usable. It’s not just usability, it’s just will they actually be able to consume it.
Trust, trust is a hugely overlooked variable in the hiring process. And I think that and I’ll go to dream, as an example. Bring a guy back up, have a big reason why this this willingness to hire failed amongst ICP2. So those fitness afterwards on the right, is because when ICP2 looks at this, they’re like, Okay, wait a minute, I’ve never really did poorly on recognition. They’re like, what is that thing? I have no idea what that does. I don’t get it. But beyond that. They were saying, wait a minute, you’re saying this can tell me all this stuff about my diet and my health and whatever? I don’t buy it. Right. This is because they were unfamiliar with how the technology can be used to fulfil those needs. So they didn’t trust it.
The other failure on trust, which we found out actually, two ways was that the branding and you kind of see it now, this is important for those of you who are involved in branding. It comes across as a startup, I guess, because the colour scheme and I know it’s Jazzy, sexy, whatever. A lot of shoppers looked at this and saw this, like, Oh, this looks like a startup, you know, how do you know the kind of things I said before the colour scheme. And it’s all sexy and everything. And shoppers were very hesitant to buy a medical device or medical devices from a startup? Because they’re like, I don’t know, you know, is it safe to use and so on and so forth. And in addition to that, because it was being perceived as a startup, they thought, well, you know, what if they pivot? Yeah, we actually had some shoppers said, Well, what if they pivot to something different? Or you know, what they get acquired? You know, I spent all this money on this thing. What then, so generate a lot of anxiety around choosing to hire the product, for example, because it feels on trust.
Do not underestimate the frequency of trust in the weight it does and making hiring decisions. I can’t tell you hundreds, hundreds of conversations I’ve had with shoppers, where we basically said you know, why did you choose Salesforce? So why did you choose this product? Or why did you buy this Nike cleats? Well, that’s what the best use. It’s actually you can google it to imitate the best.
That is it. You know, they they’re not thinking about their needs. Really, they’re not thinking they’re not calculating these others, you know, the walk in the store and they’re like, what does Ronaldo use? Okay, yeah, Nike, okay, I’ll get those. Right. Or the other one is imitate the majority. You can Google that one too. Again, it’s like well, why did you choose Salesforce? Well, everyone uses Salesforce, right? Why just use intercom? Well, I don’t all my startup friends are using intercom. Do not underestimate the weight trust has and how these heuristics around special invitation affect the hiring decision.
Next is value for money, value for money is here how the shopper determines if the cost of adoption, right is worth the time, money and effort and switching costs in comparison to other solutions. And I’ll go back to dream again. So for ICP1, once we learned, you know, we show them different packaging, different pricing, all that kind of stuff. They basically said yes to all of it. Like, yes, yes, yes, yes, yes. Because we, when they see 299, I think 299 I spend $400 a month on medication in like supplements and therapy, and they go on and on and on. So like, to them the idea that this product can do this job of helping them improve their sleep, or like, yeah, it’s a no brainer for them. But ICP2 on the other hand, I think sums it up when I this one shopper reacted to this. And he said, That’s something a rich businessman would buy. Because it was like perceived as a toy, especially with within this like lifestyle segment. You know, if you’re buying ti as a lifestyle thing. This is like an expensive toy for that rich people buy very much a luxury. Not only that, again, it was failing in so many other of the other kind of hiring heuristics. They couldn’t quite understand it. They couldn’t like figure out how relevant it was to them. They had trouble, like attaching job to be done to it. Trust was not good. And using it. Oh, yeah. Like use people anticipated saying things like, well, that’s supposed to measure me while I’m sleeping. Well, how am I supposed to get a good night’s sleep for that thing on my head? Like you’re telling me, you know, like, it’s gonna tell me I didn’t sleep? Well, well, of course, I didn’t sleep well, I had this alien thing on my head. And then He charged me three hundred bucks for that, you know, like, they just, they wouldn’t say laugh, but they just like, outright rejected it.
So again, and I kind of want to emphasise this. Simulations like this really show us. And this is what jobs theory proves us why we have to study all of this together, why you have to study willingness to hire as a ensemble process. You can’t just do one thing by itself. Like, this is why Voice of the Customer failed. I mean, it worked great. You in the 60s 70s and 80s. When you’re optimising existing products, you know, how do you make existing TVs better, you know, change to make the screen better, while it’s, you know, so on so forth. That’s kind of token but, you know, when you’re especially doing new products, new innovations, or create new markets, you know, whatever it is serious growth efforts, you got to test everything together. You know, like, for example, the Kano model, somewhat helpful, I would say, but there’s some serious issues with it. Because if you say, well, as it is, right now, if you do a condo survey will ask you all these things, but it never asked for price. You know, that’s like so if I say to someone, oh, is a bathroom in a hotel? What is that as a delight, quality performance? Or must be quality? Also, hotel must be quality of my own bathroom and hotel, give me a break. When you say, Well, what if it’s $5? A night in downtown Paris? Oh, well, okay, maybe I’ll I guess an optional, I could share a bathroom. I guess that’s okay.
So things like trust, usability, value, value for money, change the computations entirely, which again, which is why jobs three reveals, you get tested all together. And then finally, sorry, the job to be done. I left this for last as part of the hiring process because it takes a bit more explanation. So the definition going through right now is the shopper. It’s what feeds the work that the shopper anticipates a product will do to create a new state for themselves. That might seem a little up to so let me unpack that with an example. So let’s think about this one of my favourite examples, right life insurance. And I’m just gonna reduce it for the purposes of this conversation. But like, what’s Why do people hire health insurance? Well, it’s pretty, pretty clear, right? Someone has a need, they feel that well, my family’s financial situation isn’t as secure as it should be. So that’s compelling them to be open to various solutions. Maybe you’ll play the lottery, or maybe you will move to somewhere that’s cheaper, whatever has all these kind of options that they can do. And there’s one category of products life insurance that affords cash payout upon death.
The affordance cash payout upon death has to be executed in a way so that it resolves their need. So it brings them to a state of Ah, yes, now my family is financially secure. Because it doesn’t really help if you, you know, life insurance pays out a cash period of 10 bucks and that doesn’t achieve doesn’t result my need around my family isn’t as financially secure as they should be. So the key things of this so I can keep this going along, recognise that a job you’ve done is a relationship between a consumer and a product. And I’ll explain that next. The product executes the job, right? So every hire products to do the job, right, not the consumer affordances can be observable and unobservable. You know, we can observe someone listen to music, for example, it’s an affordance. But we can’t observe cash payout upon death. It’s it’s an abstraction. And also the affordance once you execute in a way which resolves the need, and of course, jobs can be for good and ill. That’s also very interesting. Why am I proposing this? Why? Why am I saying for good for ill?
Spoiler prediction, but let me let’s go back and give an example.
So what’s an example of what is the most successful product of all time? I’m not gonna do a poll, because we got to keep the conversation going on. But the most stressful probably of all time COVID-19 vaccine from from Pfizer. Last year, they made $36 billion in one year. That’s a hell of a product. So we think about well, you know, when we talk with people about, well, you know, why did or didn’t you hire the COVID vaccine, we learned what jobs they were attaching to it. So of course, we have ICP1, which a lot of people, I got this need of okay, the, I’m not as safe from COVID-19 as I should be. But then they also felt this need, I can’t do my part, if I don’t get it, I’m not doing my part. So that they look at it, they say, okay, stimulate antibodies, that’s the affordances, the word executes. And then it resolves my needs, great. But then there’s another category of ICPs, which pretty much the same thing, except they were anticipating new needs. Emerging from that, like, well, if I could get the COVID 19 vaccine, but these new needs will pop up as well, too, I’ll start becoming vulnerable to heart damage to myocarditis, or I’ll start becoming more vulnerable to as yet known problems. So this causes vaccine hesitancy, I would call it, they just were, it was not a job to done that was very appealing to them.
And the third bucket of consumers, which we found, basically didn’t feel like they have this need of like I do my part, they didn’t kind of that need was not generated in them. But then they also thought, of course, it’s kind of fringe. But some people believe this, oh, it’ll inject tracking by the government, into me. So you know, that’s an additional affordance that some people believed that the vaccine would do. And so now, you can see, now we’ve got all this stacking of these needs, new needs that are being created if I hire the vaccine. And so as a result, you get three different jobs to be done, that different consumers were attaching the product, which basically meant, you know, high likely of adopting it all the way down to basically people who would attach to this job to the COVID vaccine, basically never going to hire it no matter what.
I’m not going to go through this, but we’ve mapped out the pathway of how shoppers construct this reduced to experiments by showing stimuli in different orders, and so on, so forth, have talked about but time constraints you can get into it.
Alright, so now I got to jump to the next part here and talk about Okay, great. So now you have an idea of jobs theory tells us, okay, here’s the hiring process, and these heuristics that consumers use to determine willingness to hire. Great. So that’s the theory part. Now let’s move to actually more about how operationalize is how can I use this. So I create a method called CMA selection. And I’ll describe it as a method to forecast how target shoppers or ICPs will react to variations of a product offering. And if you saw this part of my breakout session, you have no idea what this is. We’ll talk about a little differently, though here. But if you want to look up on this, the methodology that I’m tapping into is called factorial experimentation. I will read this, you can capture this to do a screen grab or wherever you want right now to check on yourself. This is from Wikipedia, which actually describes it really well.
But the TLDR here factorial is that what you do is that for example, in A B test, your base is doing one variation. That’s it. What factorial experiment enables you to run one experiment, you can do multiple runs. And they’re called and do. I mean, now there’s like, the column like, like for x, I forgot the technical names for it. But you can do hundreds of variations. And in one experiment, and so it’s a way of designing experiments. And so how it works is you’ll choose some factor. And here, I’m having it be a factor of your offering. And you choose different levels. Levels is basically a fancy way of saying interpretation or variation. But then you look for a response variable. So so we’re looking for is we’re saying, okay, headline version two, how would that affect any of these response variables, which we’re getting from jobs theory might be, one way to think about this, is if you’re familiar with Monte Carlo simulations, you need a model, and you just the parameters, and you hit run, and you see how the different outcomes that model execute. That’s kind of what we’re doing here. With this factorial experimentation, we’re figuring out okay, you know, if I put headline version one, with affordances, 678, with pricing be whatever, whatever, you know, how does that play out? Right? Is that high risk, low risk, medium risk, whatever it is, and you conduct this experiment? And then you can very quickly discover the risk that’s in your variations or product offering.
So to give you a real life tangible example, I mean, hopefully you were there for the use case, example earlier. But if you weren’t, I’ll give the partial example here. So here, here is a offering test that we did. So it’s very similar to this, right? So we have a headline and then three variations and when to see, when shoppers react to each of these headlines, you know, how is it affecting our response variable? And I’ll just kind of jump to it. So for example, this one strategy, project manager and engagement all in one place, well, that failed pretty spectacularly on relevance and recognition. People read that and thought, I don’t get it. What do you mean, project management? What? And then they see a picture? Like, how does the thing in the like, weather do a thing on the right, like, Okay, you get some kind of like lines and graphs, I don’t get it. So right off the bat, you’re failing on relevancy or even recognition.
But also filled on trust. Because of this right here. The client thought, Hey, we should put all in one place in there, like, okay, let’s test it. And when shoppers see all in one place, they’re like, Oh, I’ve seen that before. That’s always terrible. When something tries to do one thing, or many things, it does none of them really well. So just that variation, fails on relevancy and trust. However, when we did variation to this one worked, they were able to recognise it as being new and different, they were able to also connect it to some of their needs. So it was it was relevant. And they were also think it was mostly it, alright, and they also were able to kind of start forming the job to Dawn like, Okay, I know what job, I think I have an idea what job this product does, and it’s going in the right direction for us.
So that’s just a simple example of that. And if I could really quick to wrap up here. So we have some time for q&a, at least, I will show you a little bit of how we some of the worksheets and some of the things that we have created for us. So here’s my mero board here. So we’re doing these simulations, there’s different formats that you can do. We do sometimes show shoppers, a deck, a pitch of it, which we’ll see here. Sometimes you build various landing pages, and we do something called simulated shopping, which is a little different, but we basically you know, have go to competitor websites, then go to our, the website we’re testing and see how they react and how they’re comparing and contrasting these things. But we also use these same things for email campaigns, and for advertising campaigns, for example.
So, what you end up with is a collection of stimuli of variations of your offering to predict how ICPs will will react to it. And so we formulate this into a so first to get to this. We create a our experiment plan. Pretty straightforward, right? So we got to say what factor we’re going to study with ICP, the levels in the response variable. So you would say, okay, headline, headline headline, then, you know, level one. I’m sorry, this is headline. USA headline. 100 203. Then we’re trying To make sure we’re testing for trust and recognition, for example, alright, okay, we’ve got feature number one, okay, but let’s position it this way or that way, this way, feature number two, you get the idea. And then we use that then to construct our stimuli. And then when it comes time to actually run the simulations with shoppers, we will put it into this. And those of you who murals work in here, so we actually will put into a sheet like this. And so what we’ve got here is we have the different variations that are being shown there. So you got factor one, variation one, or level one, factor two, level, factor one, level two, let me just zoom into it. Now, factor one, level one, factor one, level two. And you know, what we’re doing is, we just capture the response variables in each cell, like how each shopper each participant is reacting to the stimuli. And then what we what we’d like to do to help us make it legible at the end here is when there’s a favourable or maybe what we’ll say no risk is detected, we label a green light, it passes for example. But if we reveal a risk, ah, headline to didn’t pass trust, we’ll mark it as read. And if it’s ruined being special attention, maybe we’ll do deep read.
Ideally, over time, as you conduct these experiments, you refine it, you want to shift to a scenario where all of your responses shift from primarily red to primarily green. And hopefully, I think I’m right on time here about with a few minutes.
My answer that I propose when someone asks, How do we make growth more predictable? I’d say use job theory. And to me, it’s selection.
Subscribe for new talks, episodes, and events
Unsubscribe any time. We will never sell your email address. It is yours.
I see a great question from Dale, go for that. Can I take that?
“Great presentation. But I’m afraid I don’t understand how this is predictive.”
So it’s predictive in the sense that we can predict, or we also anticipate how an ICP will react? If so if we were to position a product in a particular way. Okay, let’s go back to dream. So the prediction is that if dream were to go to market with that product, it will tank. So that’s the prediction that we have. So that’s the prediction that we are, or one of many forecasts caught up was like forecast, because we actually predict that every permutation of it will fail. So that’s, that’s the it’s we’re saying, you know, for right now, using jobs theory, we’re saying if you put this into the market, here’s how ICPs will will react to it. And actually, so far, we find this model to work really well. In fact, in the I didn’t even know that it was working this well with Lars his team, but during the use case, on my breakout session, Lars talked about how I forgot what it was, but there was an anticipated risk. So they were predicting some sort of, I think there was a concern around the pricing model, or there was some sort of risk around the branding. I can’t remember exactly what it was. But they said like look based upon the simulations, it turns out that ICP won about 60% of our ICPs in simulation, one arm, so ICP is one in our simulations, about 60% of them have this concern around like It’s like cost overrun, for example. And so like they would reject this pricing model that was with the simulations, but ICP two would be less price sensitive, and would not reject a particular pricing model. And he says that three months later, four months later, that continues to be true, because they actually went ahead and start building that right now. And when they talk to salespeople, that ratio is true. So that’s the the predictive power of it.
Thank you. Yeah, so I’m a bit like down here. I feel a bit thick because I’m just trying to understand how that differs from or maybe it’s the same actually, things like you need artefact to test your theories, a bit like you will do in typical market research. But your framework is that in your market research, you are using five heuristics specifically, which is the one you described, and then you multiply them to try and see what happens in trajectory. That right?
Yeah, so I kind of had to rush it all together. Yeah, you’re so we have basically, there’s, you know, what we’re suggesting in jobs theory, consumers, when they’re making a decision to hire a product. There’s kind of like, you know, there’s the decision shortcuts, but they have that they’re kind of thinking about. And so when we run these tests, we’re seeing, you know, how does the stimuli affect those decision? Shortcuts? And so like in the market simulation part that I showed before, we just tried different permutations of that. So that’s hopefully I’m answering your question.
I think so. So So yeah, so in effect, in effect, you’re not predicting the, so you’re trying to predict the growth. So you’re putting something in there, not not yet on the market, because you’re testing it in the lab, in effect with a certain amount of things. But by testing these things in lab, you is gonna be is gonna give you an understanding of what the market will do.
Yes. So we are, you know, sometimes we call it a market response model, or a consumer response model. Because we are predicting or anticipating how consumers, if you were to build this thing and put it out there. We’re saying now, how they would act two years from now or one year from now, for example.
Okay. And just one last thing. Sorry, sorry to take the time. But so I came at it, maybe because the title of the presentation as in jobs to be done, and then let’s use jobs medium to predict growth. But in effect, what you’re saying is, like, less than five jobs will be done to market research a bit. Because those two things don’t exist before. And
yeah, so. I guess I mean, we think you’re, you know, I mean, to me, you know, yeah, I don’t have a great answer for you, I’m sorry. Is? Yeah. To me, it’s just like, well, you know, jobs theory is a theory. The point of every theory is prediction, if your prediction, if your theory does not predict, it’s not a theory. So, you know, we’re trying to predict, ah, will people hire or fire some product in the future? And so I’m just us figuring out how do we construct the those predictions. And every market, you know, as far as I tell, market research people in general, I don’t know, every variation of market research out there. But they’re not out there. You know, constructing using a theory, for example, to to construct a prediction. They’re just like, oh, well, here’s how people react to it. They thought it was great. Or, I mean, there might be more other things to it.
For example, you have like pricing studies, like the West Western for by Cameron, where it’s called. But yeah, there’s always different, like, typical market research stuff. But actually, jobs theory shows that those are incomplete. Because again, you can’t do pricing without also testing the product at the same time. Like, or, you know, you always have the kind of variables, right, you can’t just do needs by itself. You have to test needs at the price. That’s why That’s why Steve Ballmer laughed at the iPhone, because there’s market research had said, Oh, here’s what people want. $99 for Motorola Q. Well, that’s because historically, when they have this perception of what a smartphone is, you ask them how would you spend for a smartphone? They’ll say, Oh, $99 That’s because they never saw an iPhone before. When they saw the iPhone, they’re like, oh my gosh, y’all spend 500 bucks on that. And then look to now it’s $1,300!
Alan has dedicated himself to making predictions on which growth strategies will and won’t work, and creating alignment via a shared language of growth.
His background is in product and growth, working in startups big and small for 15 years. He founded Revealed, a product, growth, and go-to-market strategy firm, in 2015. He works with clients including Twitter, Google, Pipedrive, Asana, American Express, Lego, Arlo, Hubspot, Netgear, Dreem, and Backless Basics.
He’s the author of ‘When Coffee and Kale Compete’ and is currently writing a second book on JTBD, ‘Jobs to be Done. Understanding Needs. Predicting Adoption.
Subscribe for new talks, episodes, and events
Unsubscribe any time. We will never sell your email address. It is yours.