Speedrunning the Skill of Demand

This is Part 1 in a series of essays about Customer Demand.

Let’s talk about demand.

What I mean when we talk about ‘demand’ is customer demand — that invisible thing that makes business possible. Skill at understanding demand is what underpins the domain of marketing, sales, and product. There’s an old Peter Drucker saying that goes “the purpose of a business is to create and keep a customer” — which is actually quite profound if you pause to think about it. Without a customer, you have no business. In order to create a customer, you must find and then meet an unmet demand. Most things in business are ultimately in service of this.

Demand is one of the first things you learn when you go into business for yourself. People will talk about it obliquely, though. They will say things like “get out of the building and talk to customers”, or “sell painkillers, not vitamins” or “build something people want”. Sometimes they will say “first time founders talk about product, second time founders talk about distribution.” At the end of the day, all of these quips are about demand.

I learnt about demand the way most people do: the hard way. I did my first startup in my freshman year in university. It died a pathetic death two years later: I had not built something people want. By the time I left to go into business for myself — seven years later — I had built and launched six products with varying degrees of success. The last two products I launched was part of a bootstrapped company, one that I built as the boss’s ‘right hand man’; if I were to guess, I would say that those two pieces of software have made several million dollars worth of profit for him and for my old team over the past eight years.

But still, I was not good at demand. You can luck into building saleable products — a business mentor pointed out to me that the product category we worked on (point of sale systems) was a relatively kind environment for product development. He was right. And so I decided to get good after I left the company.

It’s a little wild that we’ve never talked about demand here at Commoncog. There are thousands of articles about understanding customers. There are so, so many frameworks. Pretty much every good marketer and sales rep I know is obsessed with demand — even if they might not talk about it explicitly. Every attempt at new product innovation is ultimately concerned with seeking out latent market need. Hell, what else do early-stage founders obsess about except for product-market-fit?

So why are we talking about it now?

There are two reasons.

The first is prosaic. At the end of 2024, we at Commoncog completed our very first application of the Jobs to be Done (JTBD) Framework — an effort that took us around five months. We’re starting to apply the findings of that framework now. I think it would be useful to write about that experience, when it is still fresh, from the perspective of someone who struggled to put it to practice. JTBD is a lot harder than it looks.

Second, I was introduced to a book named The Heart of Innovation late last year, when our JTBD interview process was still ongoing. At the time, this was a fairly new book, with a terribly bland name and a generic cover. The contents of the book, however, are remarkable. I believe The Heart of Innovation is the first new contribution to the body of knowledge we have around customer demand since the JTBD framework — which itself was developed in the early 2000s.

I want to summarise the book’s ideas. But in order to talk usefully about the book, we should go over the state of the art: what do we currently know about identifying and hunting down customer demand? How did I speed-run this skill, when attempting to learn it for myself?

In this series, I’m going to present a highly compressed, somewhat idiosyncratic introduction to the topic. This will not be a long series, because much has been written about demand elsewhere. If you’re in sales, or in marketing, or if you’ve attempted to create and sell a product or service of your own, it’s highly likely that you know something about demand already. I hope you’ll give this series a shot, at least until we hit the book summary of The Heart of Innovation.

(Alternatively, you can skip over all this and go out and buy a copy of the book now — I guarantee you’ll learn something new).

I was forced to get good at understanding demand in the years between 2018 and 2021. I knew I was finally decent when I doubled the annual recurring revenue of a software company at the end of 2021, adding some seven figures to their top-line with no additional marketing spend. I intend to summarise everything I’ve learnt during the period. Along the way, we’re going to publish some new cases to illustrate these ideas; all of these cases will live in the Idea Maze concept sequence.

And then we’ll talk about the revolutionary new ideas that The Heart of Innovation introduces.

Let’s get started.

How I Learnt Demand

I learnt demand in a very obvious way: I was bad at sales and marketing. When I left the point of sales business, I knew I was good at operations and at running small software teams; I was horrible at the customer-facing parts of the business. What few client interactions I’d done under my old boss had turned out badly. I saw my boss at work with new customers; I also met with our best salesperson once a month or so. I knew they understood something I didn’t. And so when I left the company, I knew that if I was going to be effective at business, I would need to get good at one of either sales or marketing, and quickly. I picked marketing. And then I started Commoncog.

Starting a blog was a pragmatic decision. You can think of marketing as having two broad categories: paid or unpaid. Paid marketing was the game of buying ads on various platforms like Facebook and Google. This is a skill of its own; I knew folks who were good at this. Unpaid marketing was basically content marketing (producing videos, podcasts or articles) that would gain word-of-mouth distribution and eventually — hopefully! — influence a buying decision. Unpaid marketing also included Search Engine Optimisation (SEO). I had no product to sell, and no money to buy ads. So unpaid marketing it was: I started a blog to get my hands dirty, and used the experience to get better at both content marketing and then (much later) at SEO.

Most folks talk about ‘getting good at marketing’ in terms of learning marketing skill, not in terms of learning demand. It takes awhile before you learn that ‘understanding demand’ is a related skill that underpins the skill of marketing (and sales, and product).

In his 2022 book Learning to Build, entrepreneur and JTBD framework creator Bob Moesta argues that great product builders are really good at five different skills, which he’s observed again and again across the many hundreds of companies and clients he’s worked with. One of those skills he calls ‘uncovering demand’:

Uncovering Demand: people with this skill can detach from their product or service and look at the demand-side of the equation — seeing struggles, context, and outcomes. (emphasis mine)

That detachment turns out to be important. Folks who are good at understanding demand will think about the nature of customer need without considering the constraints of the company or the product they’re currently involved with. This is much harder to do than you think, especially if it’s your company, or your product that you are thinking about. Paradoxically, this detachment is what helps them build better products.

What is true for great product folks is also true for marketers. Good marketers think in terms of tactics. Great marketers go further: they know all the tactics that the good marketers know, but they also have an ability to uncover and understand the latent demand that is separate from their product.

Perhaps a concrete example would be useful here. When you do content marketing, you quickly learn that you need to do three things:

1. You need to pick a topic that is interesting to your target customer.
2. You need a differentiated take on the topic — that is, a take that is different from other publications covering the same topic.
3. Finally, you need the topic to be large enough, so that you don’t run out of things to write about.

The central conceit of content marketing is that if you write sufficiently interesting things, on topics that your target customer is interested in reading (or is interested in searching for; SEO is a valid variant of this), they will hang around and eventually think about purchasing your product when they finally feel the need.

While all three elements are — to some degree — all important, that first thing is probably the more important consideration for a successful content marketing strategy. It doesn’t matter how attention-grabbing your content is if you’re not attracting folks who would someday buy your product. If you are doing content marketing for a business intelligence tool (think: Tableau), the first step to effective content marketing is less “plan out a content schedule” and more “figure out who buys Tableau, and then figure out what they’re interested in reading about”.

But this is merely level one of the skill, and not really an interesting level. A more effective level thinks about demand in a more nuanced way. Let’s pretend that you are the head of marketing for Tableau. Being good at uncovering demand looks something like this:

• First, you should figure out how a purchasing decision is made for something like Tableau. You listen to sales calls, preferably a double digit number of hours. You stop when you can predict how a given call will go.
• You dig deep into the properties of the most successful customers using Tableau. These are customers that you want to see more of. Usually this requires you sit down with folks from the sales and customer success teams, and it requires you to look up numbers for sales pipeline, retention cohorts, and customer support tickets. You want to find leads that close quickly, stick around for a long time, and give you relatively little customer support pain.
• Next, you figure out how to reach customers with those properties. If you are a content marketer, this will look like asking: “now that I know my target audience (i.e. folks with similar properties to our best customers) what do they read?” And then: “let’s plot a content schedule around that.”
• That’s all well and good, but if you think of yourself as an an effective marketer, not just a content marketer, you would say to yourself: “Hmm, interesting. Maybe these folks don’t read. Never mind. Where else can I find them?” This might be specific conferences, or ads in airport terminals, or LinkedIn cold outreach, or specific types of podcasts. You can then spin up tightly scoped initiatives — with small budgets! — for those channels.
• And if you are even more effective, you will think “huh, it seems like folks are buying Tableau because they need pretty visualisations for ad-hoc requests like board decks. Tableau also gets used a lot when someone needs to impress with pretty-looking (but actually ineffective-for-decision-making) graphs. Meanwhile, our customers churn out of Tableau when they start using it for operational data reporting. We should redesign the product and the sales process to emphasise the former, and discourage the latter. Let me go talk to product and sales with my findings, and influence them to move in that direction …”

(And perhaps now you have an idea for how I could start out as a part time content marketer and then generate over a million in additional annual recurring revenue, over an eight month period, for a software company that was in the same data visualisation space.)

But I digress. The key thing to notice in that sequence of questions is the relentless focus on finding and then fulfilling customer demand, in a way that ignores neat categories like ‘marketing’ or ‘sales’ or ‘product’. A good content writer may think about things like craft and topic selection and sentence structure. A good marketer thinks about the customer. And the key thing to internalise is that the customer does not care about your writing, or about your product. A customer does not buy your product because they like your product. A customer buys your product because they believe it will turn them into a more awesome version of themselves.

This sounds very trite but I assure you that it isn’t. I’ve spent lots of time with product owners and founders who continually think from the perspective of their product, their baby, and their company, and not at all from the perspective of a customer wanting to change their lives. Often the difference is subtle. There are lots of little cues when you’re reading marketing copy that will tell you if a company understands who it is serving. The ones who aren’t great will typically stray — just a smidge — on the side of navel-gazing.

So, to recap: the job of marketing is to figure out customer desire, attract customers with this desire, and then frame the product as “this thing will make you really awesome.”

It took me awhile to learn all of this.

Sales Safari as a Method to Get Good Quickly

How did I learn this way of thinking? Hell, how do most people learn this style of thinking? I suspect most folks learn through osmosis. I certainly did, at least at first. In the beginning I spent a large amount of time listening to marketing experts talk shop on podcasts (a technique I wrote about in To Get Good, Go After The Metagame), and eventually came to the realisation that most good marketers — and entrepreneurs! — understood demand at this fundamental level.

The problem with osmosis, of course, is that it is really inefficient. No, looking back, I got better faster only after I learnt Sales Safari from Amy Hoy and Alex Hillman. I paid for their flagship 30x500 course back in 2019, which was really expensive. Thankfully, they now offer a cheaper program called Sales Safari 101, which focuses only on the demand discovery technique that they developed. (30x500 focuses on the full lifecycle of creating a new product from scratch — Sales Safari is taught in the first module of that course). A few years later, I realised that Hoy and Hillman had invented something remarkable: they had synthesised years of business experience and several dozen books on sales and marketing down to a single technique; much of the magic of 30x500 was really hours of practice at uncovering demand.

The technique has been so valuable to me that I now use it as an accelerated training tool: it is an intense, but extremely powerful method to get someone who is a novice to become good at understanding demand. I’ve found that it is much easier to teach marketing skills once folks have gone through one complete Sales Safari project (which lasts between two to three months).

What is Sales Safari?

Sales Safari is an ethnographic technique to understand potential customers. In the original 30x500, Hoy argues that you should not talk to your customer, but instead use Sales Safari on forum threads and social media hangouts — places she called ‘watering holes’. This was by design: most folks aren’t great at talking to potential users, and tend to get mixed signals from the process (witness the novice founder’s problem of “ahh, I showed my product to the user and she seemed excited — but two months later it doesn’t seem like she’s going to buy.”) Hoy’s process sidestepped this challenge. You can’t lie to yourself if you’re not talking to customers; instead, you are forced to observe customers in the wild. You want to watch potential customers complaining to each other or shooting the shit in a forum or Facebook Group, or subreddit. Observing customers in such ‘wild’ habitats is as pure a signal as you’re going to get.

Of course, once I mastered this technique, I started applying it to sales calls and support calls and in user interviews. I would often find myself listening to a call whilst cooking or running, pausing to jot down some note according to the Sales Safari template.

How does Sales Safari actually work? I’ll describe the technique in full, but I should warn you that my descriptions are not going to be enough. Over the past five years, I’ve taught Sales Safari to enough folks to know that you really need to practice this for a month or two in order to get good. More importantly, you need feedback from someone who is already good at Sales Safari in order to progress rapidly. The way I typically train people is to let them present their notes before I do, and then iterate until they finally catch most of the things that I see. Initially, novices are surprised at how much they miss. As they get better, our notes will converge.

All of this is to say that you should buy Hoy and Hillman’s course if you want to get good. They are good teachers. My only caveat is that you must put in the work. Sales Safari is more difficult than you might think, especially if you’ve never done ethnographic research.

The Sales Safari Technique

Take a piece of paper, or open a new document in your favourite note-taking app.

Write the following headings:

1. Rough notes — Everything that doesn’t fit into the other categories goes here.
2. Pains — Here you will list all the pains that folks allude to or talk about during the sales call (or in the forum thread).
3. Worldview — Worldviews are approaches or stances towards the world. This is one of the trickiest element of Sales Safari, but also one of the most important. For instance, Android users have the worldview “my phone should serve me, and give me the freedom to do what I want.” iPhone users have the worldview “my phone (and my devices) should Just Work, without my messing with finicky settings.” Users will rarely talk explicitly about their worldviews, so much of the worldview training in Sales Safari is directed at uncovering worldviews through keen observation.
4. Jargon — Jargon is a list of any non-standard English word used during the call or forum thread (or any standard English word used in an unusual, domain-specific way). Just keep a list.
5. Recommendations — When people talk about pains, they tend to also talk about solutions to those pains. If you’re doing Sales Safari on a forum thread, it’s likely some of the commenters will come in with recommendations for solutions. List them under this heading.

Your job is to listen to a customer call or go through a forum thread and then fill in the various headings in your document.

It will take you more than an hour, on average, to process one call or one decently active forum topic using Sales Safari. You will feel exhausted after that hour. This is normal, and it means you can only do so many Sales Safari analyses per day.

Why is this so? The thing that makes Sales Safari difficult is that so much of the analysis is about what is unsaid. I remember one session where we were listening to a customer support call. The person our rep was talking was the Head of Data for a multi-million dollar, growth stage e-commerce startup. He spent a full 20 minutes, on a 60 minute call, ranting on and on about his BigQuery bill. BigQuery is a cloud data warehouse provided by Google; it charges according to usage. What had happened was that one of his analysts had spent a few hours spelunking in a BigQuery-hosted dataset the month before, using our data visualisation tool; at the end of the month, they discovered that the analyst had racked up $10,000 in costs from that single day alone.

“This is unacceptable,” he said, repeatedly, “Absolutely unacceptable. This is too risky. If this can happen once, this can happen again. Terrible! How do people handle this? Is there a way that we can track our BigQuery costs live, in your user interface?” He went on like that for several minutes.

In our review session, the marketing folks I was training listed “Finds BigQuery costs expensive” as their top pain. Their notes looked something like this:

Raw Notes

• Name XXX XXX, LinkedIn profile: <link to LinkedIn>
• Head of Data of XXX, a X00M startup. Runs a team of 5 data engineers and 5 data analysts.
• Custom data reporting stack, built over two years, on top of commercial, licensed OLAP database. Analysts have to write raw Javascript to create new dashboards.

Pains
Finds BigQuery very expensive.

• Analyst spent a few hours month before running queries in our tool on top of BigQuery, cost additional $10k
• “This is very worrying. This is unacceptable. Is there a way we can keep track of our BigQuery costs in your software UI?” (21:34)

… Other Pains, Worldview, Jargon, Recommendations, etc snipped …

(Note: all assertions about the customer should be backed up with excerpts from the call transcript, with those excerpts put in quotes, and with the exact timestamp attached. I’ve found that this makes it easier to replay the call and review the source material with precise timestamps if necessary during the analysis process. Sometimes this is because you want to evaluate the level of the pain in the customer’s voice.)

I stopped the discussion and pointed out that this assessment didn’t make sense. “Consider the context,” I said, “This guy has spent two years building a custom data reporting stack. The license for the commercial OLAP database they use probably costs in the tens of thousands each year. He runs a team of five data engineers and five data analysts. The data engineers build the company’s dashboards from scratch, using Python and Javascript. Assuming $100k per person per year, that’s $1M in annual personnel costs right there. (Yes, many of these folks are in Eastern Europe, where things are cheaper, but let’s just round up). $10k in BigQuery costs is nothing for this team.”

No, I argued, the key pain this Head of Data was feeling was likely that BigQuery costs were unpredictable. That is a very different pain from “BigQuery is too expensive”. It implied a different set of features and solutions. I told my team that when you are doing Sales Safari, you need to take the full context into account — both what is said and unsaid. Also: notice that we did not talk about our product at any point in this process — we were primarily focused on uncovering latent demand, in whatever form that might take.

This sort of analytical discussion repeated again and again until our notes converged — an arduous process. If I remember correctly, we did three customer calls per week for the first two weeks, and then split the workload for the next two weeks when I was certain they were good enough, for a total of 11 customers. This took us a whole month, at about an hour per call. We were exhausted by the end of it.

All of this is what makes Sales Safari difficult.

Using your Sales Safari Output

The end result of a Sales Safari research project is a folder full of notes. These notes are hard-won ‘Customer Gold’, as Hoy and Hillman call it. They form the basis of all follow-on marketing and repositioning activities.

Once you start using Sales Safari in your follow-on activities, the note structure begins to really show their value.

• If you are conducting a Sales Safari exercise with best fit customers, the Pain section of your notes will tell you what these customers actually bought you for. You can lump customers into categories based on this analysis, and then opt to reposition around one or two of these categories. Repositioning means setting a new frame and direction for product, marketing, sales, and so on; how to do this is the topic of an amazing book called Obviously Awesome by April Dunford.
• On the other hand, if the subject of your Sales Safari is a forum thread, the Pain section will be a collection of pain points you may start building products or writing copy for.
• The Worldview section helps with targeting customers. As a marketer, it is much easier to target worldviews, as opposed to targeting demographics. This is actually some savvy slight-of-hand by Hoy and Hillman: older marketing frameworks tend to focus on demographics (e.g. ‘hispanic office workers between the ages of 30-40’). The focus on segmentation through customer worldview is a comparatively new marketing approach. It is also more effective. People don’t buy because they are of a specific race, age, or profession; they buy because they see the world in a certain way, and your product helps them make progress towards their goals given a certain worldview and a certain set of pains.
• The Jargon list comes in handy when you have to write copy. You know the old advice to “use language that your customers use”? The Jargon section helps when you need to write something customer-facing — be it marketing page or sales deck. If you have a Sales Safari stash, you may simply flip through a number of research notes, scan through the Jargon sections, and then use the lists as the basis for word choice.
• Finally, the Recommendations section is useful when you have to write content marketing, or when you have to think about possible alternatives to your solution. Folks tend to have common recommendations for pain points that they face. The Recommendations pile in your Sales Safari research is essentially a resource that you may use to generate ‘this is why our product/solution is better than X’, where X is a recommendation by a potential user. Recommendations from your ideal customer may also illuminate how folks in your target audience think about the problem they want solved — “buying a competing product” is a different solution from “hiring an intern to do this tedious job for me” — and that tells you something about the real competition your product faces in the market.

In my experience, running one Sales Safari project — which takes anywhere between two and three months — is enough to teach the basics of demand to a sufficiently bright, sufficiently motivated person. I have trained at least two people who have gone on to good careers with Sales Safari. Their understanding of demand has formed the foundation of the rest of their skillset.

In other words, it is easier to layer on new skills — SEO, growth, sales enablement, email campaigns, newsletters, copywriting — once you understand demand. My hope is that this essay helps you see why.

Wrapping Up

Let’s wrap up.

This piece serves as the kick-off for a mini-series on understanding demand. I opened with an argument for getting good at demand, as distinct from leaning higher-order skills like sales, or product, or marketing. I then told you a brief story of how I acquired this skill, and introduced you to the technique that I used to accelerate skill acquisition — both for myself, and then for a few others.

I’ve also alluded to some of the results that you’re able to get if you apply Sales Safari strategically — something that we’ll talk about in a future instalment. I was joking with one of my teammates before writing this essay: perhaps I am overselling myself on this topic. But then I caught myself — it is not normal to make such a material impact on revenue for multiple businesses, nor train multiple folks who then go on to become good marketers or founders. Understanding demand is potent. Sales Safari is the fastest way I know to get someone to a good understanding.

One neat thing about Sales Safari is that it packages a surprising number of effective demand concepts into one simple technique. Notice that we have — over the course of merely discussing the form of the method — covered the following ideas:

• “When a customer buys your product, they don’t actually care about the product. They care about themselves. At the point of purchase, a product is merely the thing that they believe will make them a more awesome version of themselves.” This is the core idea behind Kathy Sierra’s 2015 book Badass: Making Users Awesome. Hoy and Hillman simply repackaged this insight in a way that is useful.
• “The customer purchases your product to solve a pain point.” This is another idea that seems obvious; it is the premise for everything that we’ve covered in this essay. A huge chunk of Sales Safari is to uncover audience pains that you may target — in your marketing, through your product, or perhaps through a service you can sell.
• “Not every customer will purchase your product to solve a pain point. Only customers with a certain worldview will do so. Your job is to seek out that worldview, so that you may filter for it.” In marketing terminology this is known as ‘segmentation’. It’s typically the case that customers will gravitate towards certain solutions and not others in response to their pains. (iPhone users are different from Android users because they value different things). Your job is to find customers who are a good fit for you. Segmenting customers according to worldview is more effective than segmenting according to demographics. Your job is to isolate the worldview that works best for you, articulate it well, and then use it to find (and select for) customers who share that worldview.
• “Use words that your customers use to describe your problem or solution.” This one is obvious.
• “The job of sales and marketing is to seek out and then filter customers who are the best fit for your product.” Why? Well, because the job of a business is to create a customer.

In the next instalment, we will build on some of the foundational ideas covered in this piece. What happens when you want to go one step beyond identifying pain? What happens when you want to bend the arc of the buying decision?

This was exactly the problem that Bob Moesta — alongside legendary business professor Clay Christensen — sought to solve with the Jobs To Be Done framework. They came up with the framework in the early 2000s. We’ll look at that next.

Follow Ups

• Watch this demo video of Sales Safari, conducted by Amy Hoy at La Conf Paris 2013.
• Check out 30x500, their product development course, or Sales Safari 101, the cheaper, more tightly scoped course on the demand discovery technique.
• Kathy Sierra’s Badass: Making Users Awesome.

Originally published 14 March 2025, last updated 14 March 2025.