AI In Sales & The Jevons Paradox

A month or so ago, I wrote about the emerging AI business model and its potential impact on white-collar employment. As I said in the post, I'm skeptical that AI will reduce the number of white-collar workers, particularly salespeople. With that in mind, I came across a concept called the Jevons Paradox. 

The Jevons Paradox says that when technological improvements increase the efficiency of a resource, that leads to higher consumption of that resource, not less.

A couple of examples:

Technologies like home insulation, in theory, would reduce the need for heating and cooling energy consumption. But the effect was that people invested in larger homes resulting in more overall energy consumption.

Electric/hybrid cars bring down energy costs, leading to more frequent driving and longer trips, and more overall energy consumption.

This concept also applies to work.

Flexible work tools like Zoom have added convenience but also made it possible to work more, increasing the number of hours worked. 

Communication tools like Slack and email make communication more efficient but have also led to far more communications and increased workloads. 

I think we'll see the Jevons Paradox at work with AI in sales. As AI takes on more of the job of a salesperson and does things like lead scoring, opportunity prioritization, CRM automation, communication personalization, sentiment analysis, etc., it will bring down the cost of sales. That means that each salesperson costs less per dollar in sales than they did prior to the adoption of AI. That will cause companies to want to do more sales and hire more salespeople to do the work.

The myth is that automation makes the worker superfluous. In reality, the opposite is true because it’s an opportunity for the ambitious to do more. Marketing automation is a great example. Facebook Ads can effectively run highly targeted, comprehensive marketing campaigns, in theory reducing the need for digital marketers. When, in fact, those platforms made marketing cheaper, which led to major increases in the need for more digital marketers to manage these programs and to do more marketing. The same will be true with AI in sales.

When something gets cheaper, we tend to do more of it.

AI & Sales

A few weeks ago, Tomas Tunguz from Theory Ventures gave an insightful talk on the state of SaaS Go-to-Market. One key takeaway was the explosive growth of AI use among sales teams. This isn’t surprising. Sales teams, by nature, are quick to adopt new tools because they’re profit centers. Anything that works will be adopted quickly because it impacts the top line. AI is helping sales teams in several areas—lead scoring, customer research, training, follow-up automation, competitive intelligence, and price optimization, to name a few.

Sales leaders are reporting lots of efficiency gains thanks to AI, but despite the boost in productivity, AI hasn’t moved the needle when it comes to deal conversions or bookings growth.

The reality today is that AI in sales — as it is in many functions — is really like having a bunch of productive interns. They can handle lots of grunt work, but their output needs to be checked, they’re tough to train, they lack company and situational context, and they’re not equipped to make important decisions. AI can’t yet do what great salespeople do. It can’t build emotional connections, navigate complex negotiations, leverage intuition, adapt in real-time to low-information situations, or respond to subtle cues or nuanced human emotions surrounded in layers of context. In short, AI is really good when it has lots of great training data to work with, which is the polar opposite of what a good salesperson does: they shine and show their real value when they lack data, and the next step is unclear, and they have to jump over hurdles to acquire more information or rely on their emotional intelligence, intuition, instincts, and human connection to make a good decision without it.

That said, I do believe the efficiency gains AI provides will eventually translate into better conversions and growth. But I’m skeptical how high up the chain it can go in terms of high value sales activity.

AI in sales will be a fun one to watch. Sales results are pretty measurable, and teams will quickly adopt what works. We’ll see.

Why Is It So Hard To Sell To Health Systems?

As a seller, you want to sell to an organization with a single decision-maker who’s free from regulatory constraints, has high margins, a risk-taking mindset, plenty of cash, and no legacy systems in place to slow things down. This is the exact opposite of a typical health system.

Many salespeople claim their market segment is the toughest. But those who sell to health systems might just take the prize. Health systems are notoriously challenging to sell into. To sell to these organizations more effectively, it’s helpful to consider why that is.

1. Complex Organizational Structure

Health systems have very complex ownership and decision-making structures. There are multiple layers of authority, often with competing interests, incentives, and priorities: executive leadership, service line leadership (orthopedics, cardiology, nephrology, etc.), procurement leadership, and technology leadership, all of whom can veto a decision. Clinical needs can also quickly shift priorities, sometimes sidelining other initiatives. And then there’s the troublesome “innovation” team, often connected to a venture investment fund with its own set of disjointed incentives. Selling a product to a health system often requires buy-in, alignment, and approval from all of these stakeholders. A key part of the job for a salesperson is to help the buyer navigate their own internal processes to enable a purchase.

2. Regulatory and Compliance Constraints

Health systems face numerous layers of regulatory oversight that can significantly slow down procurements. Requirements include HIPAA, accreditation standards, state licensing, anti-kickback laws, and CMS guidelines, to name a few. On the technology side alone, there are the HITECH Act, EHR incentive programs, FDA software regulations, FISMA, and interoperability standards. Vendors must undergo rigorous vetting and compliance processes, which can radically slow and often end a sales cycle. It’s crucial that sellers have a deep understanding of the regulatory and compliance issues that overlap with their products.

3. The Dominance of the EHR

Few industries have a software vendor that wields as much influence over internal decision-making as EHR vendors do in healthcare. These are giant, long-term contracts, and depending on where a health system is in its EHR implementation, dozens to hundreds of EHR employees can be embedded within the health system. Often, if you’re selling a software product, someone in the health system will consult their EHR vendor before even considering a conversation. These EHR vendors frequently have ancillary or directly competing products, and there’s a saying among health tech salespeople that the large EHR Account Managers are trained to respond to any new software procurement by saying, “We already do that.” And often, once you get approval to move ahead, a lengthy integration conversation must be waded through, both in terms of feasibility and timing.

4. Low Risk Tolerance

Given the life-and-death nature of their work, inherent low margins, litigious concerns, mostly not-for-profit structures, and heavy regulatory burdens, health systems rightfully prioritize avoiding mistakes over taking risks. This makes selling a new, unproven product quite challenging. They want free trials, extensive case studies, flexible contract terms, and numerous customer references. Their ROI standards are stringent, and they rely heavily on evidence-based data before making a decision. Also consultants are often brought into the process as well to help reduce risk and add rigor, adding further scrutiny and potentially misaligned incentives. A healthy cautiousness is a key part of the culture of a health system.

5. Health Systems Are (Mostly) Local

Over time, most industries will consolidate down to a few national or global players. This isn’t true in healthcare. There are 2 to 4 dominant health systems in each large metropolitan area in the country. There is some overlap via large regional players or national health systems, but not a ton; it’s a highly fragmented space. There are a bunch of very large health systems businesses scattered across the US. There are 5 US airlines whose annual revenue exceeds $5 billion. There are nearly 10 times as many US health systems that meet that threshold. These factors make selling more difficult for a variety of reasons. You don't have a consistent set of priorities across markets to sell into. There's often allegiance to local or regional vendors. It's hard to leverage competition because a New York health system doesn’t care much about that big account you won in Houston. More broadly, the lack of scale in these local health systems trims margins and makes budgets much tighter than they might be on a national scale. 

6. Supply & Demand Mismatch

Over the last decade, more than $100 billion in venture capital has poured into health tech startups, creating a vast supply of vendors that need to show a return on that capital by selling their products to health systems. The supply of products far exceeds the demand, leading to vendor overload for health systems. There’s a lot of noise out there that buyers are forced to wade through. Many health systems have responded by launching innovation teams, some with venture funds attached, which attract vendors to test products using health system staff and provide the fund with deal flow. These innovation teams often wield outsized influence, creating incentive misalignment and internal conflict. Even when these aren't in place, you can bet that the health system you're selling to is pretty sophisticated at sidestepping salespeople.

Asking Great Questions

Pushing oneself to ask great questions is extremely important. It forces you to think critically, accelerates your learning, improves decision-making, and helps build trusting relationships. Here are some things that have helped me ask better questions:

Tell yourself a story and use questions to fill in the gaps.

When I'm interviewing someone, being interviewed, or just having a conversation with a colleague or founder about a business problem, I typically take what the other person is saying and try to tell a story that makes sense in my mind. If someone is explaining a new company they'd like to start, I'm listening, but as I hear them talk, questions pop into my head about things I want to know more about or inconsistencies with my preconceptions. This is the trigger for most of the questions I ask, and it’s a useful way to generate dozens of questions in a single conversation.

Of course, you have to temper your questions and prioritize the most important ones so you don't sound like a nut. But forcing yourself to tell your own story based on what you're hearing—and fitting it into your worldview, or using it to learn something new and adjust your worldview—is a great way to ensure you're asking high-value questions.

Be ruthlessly authentic and ask “dumb” questions.

Have you ever noticed that whenever you ask a “dumb” question, it often turns out to be a great one? Dumb questions are wonderful. They typically simplify the topic, challenge assumptions, align the conversation at a high level, and enhance inclusivity. The trick is having the humility and courage to risk showing that you don’t know something that others do. This is definitely a risk worth taking. Try to never hesitate to do this. Be authentic. The best questions are about the things you really want to know. As someone once said, “Not knowing is not ignorance; not seeking to know is.”

Ask what people think, feel, or would do—not what they know.

I almost always find that opinions based on facts are way more interesting than simple facts. Instead of asking a job candidate, “What is our competitor's product strategy?” try asking, “What do you think of our competitor's product strategy?” Similarly, don’t ask your real estate broker, “How can I get a better price?” Instead, ask, “How would you go about getting a better price?”

This tactic encourages the person to really think about the answer and provides you with genuine insights rather than just reciting what they know. Also, give them time to think and express their full answer. Rushing them only leads to less interesting responses.

Bookings As A Lagging Indicator

Bookings (the value of contracts signed within a specific period) is a crucial metric for companies to watch. Investors watch this number very, very closely. Boards will put enormous emphasis on it. When a deal is booked, the product then gets delivered to the customer, which turns the booking into revenue generated in a specific period, which equates to the top-line growth of the company. Bookings are the tip of the spear. It’s a leading indicator for revenue.

Investors will also look closely at qualified pipeline (the pool of potential sales opportunities that are deemed highly likely to convert into bookings) as that is a leading indicator of bookings.

Bookings and qualified pipeline are watched closely and are heavily scrutinized.

The problem with placing too much focus on these numbers is that a sales and marketing team is limited in how much they can move these numbers one way or another in a specific period. If a company crushes their bookings in a period, it generally means that there was a bluebird deal or that goals weren’t set accurately or that there was an external macro event that caused a large swing. Rarely are sales and marketing teams able to swing these numbers up and beyond expectations in a major way. The reason is that bookings are capped by the TAM (total addressable market) or, more specifically, SAM (serviceable addressable market) available to them. I wrote about TAM, SAM, and SOM a few years ago, find that post here. So, the reality is that while sales and marketing teams can do great things, they are limited by the stuff they have in their proverbial bag that they can sell. To really move these numbers and continue to grow, companies need to create new SAM at a high rate. So, while pipeline is a leading indicator for bookings, SAM creation is a leading metric for pipeline.

To make this point more concise, bookings growth is dependent on product investment decisions that were made 1, 2, 3, or even 5 years ago.

So, while it’s obvious that companies should be focused on in-period bookings and retention and profitability metrics, arguably it’s more important for companies to be focused on in-period SAM creation such that the cap on bookings growth in future periods gets higher and higher. The reason this is arguably more important is that product investments made now can drive far larger swings in growth in future periods than a sales and marketing team can in the current period. If two years ago a company made large investments in new products and new SAM creation, bookings will be high in the future. If two years ago they made no investments in new products and new SAM creation then bookings will be low in the future.

So, while obviously investors should be asking companies how bookings are going in a specific period, they’re really looking at a lagging indicator for good or bad investment decisions that were made in the past. They should place equal emphasis (arguably more) on how much new SAM is being created in that same period, as that’s the number that’s going to drive material and sustainable growth.

Incentives, incentives, Incentives

"Show me the incentive and I will show you the outcome" - Charlie Munger

I’ve found that when you think understanding incentives is really, really important, you still don’t understand how important it is. When you find yourself in a difficult conversation or challenging situation with an employee, a customer, a partner, or a vendor, pause and make sure you understand the incentives of all the stakeholders. It almost always gives you instant clarity and gets you unstuck.

Good Conversations & Good Selling

As a sales leader, I’ve often told my teams that, while I want to close every deal we can, it’s perfectly fine if we don’t, as long as we understand exactly why the buyer didn’t buy.

As a seller, I used to have this mindset during sales conversations: I wanted to be sure that if the buyer didn’t buy, I could explain why in great detail to anyone who asked.

The trick here is that forcing yourself to understand why a buyer didn’t buy also forces the right sales behaviors. You have to do a bunch of good things to be able to do this well. You have to:

• Talk to the right people.

• Understand their role.

• Understand their decision/buying process.

• Understand their incentives and priorities.

• Understand their problems

• Understand what other solutions they have.

• Understand what they like/don’t like about the product.

• Understand how they view the competition.

• Understand how they think about ROI for the product.

You can’t understand these things if you’re not having good conversations. And good conversations are the thing that drives sales. Sellers should be less interested in making the sale and more interested in deeply understanding the mindset of the decision maker, regardless of the ultimate decision. This mindset, repeated over time, leads to better and better high-quality conversations. And better conversations = higher win rates.

Beware Of The Warm Intro

Jen Abel from Jellyfish posted this great Tweet the other day.

When a friend or colleague makes an intro, you feel somewhat burdened to take the call, to be nice, and to make it seem like the intro makes sense and that there's a deal there. But so often, this isn't the case. Sure, the person making the intro has the best of intentions in mind, but often, they have no idea whether a partnership between the two organizations being introduced actually makes sense. And they're often doing it to build up their own social capital in the hopes of leveraging that capital for themselves down the road. This is ok and sort of the way humans work, but it's crucial not to get fooled by the false momentum that comes from these connections. This is particularly difficult when the intro is centered around a more ambiguous partnership or when it comes from someone more senior than you. It takes real discipline to drive clarity and prioritize the opportunity appropriately.

Warm intros should be qualified and disqualified using exactly the same criteria as a cold outreach. Don't get caught up spending time with someone who is happy to take your call but ultimately doesn't have a high-priority problem that you can solve. 

Two Rules For Negotiating

There are two rules for negotiating that are absolutely essential.

1/ Be able to make the other side's argument as effective or (ideally) more effective than they can. If you can't do this, then you’re lacking empathy for the other side which is very dangerous. If you can’t do this, then you likely don't truly understand the issues at hand, which means you won't be as effective as you could be. Never negotiate without doing this. 

2/ Internalize your BATNA (best alternative to a negotiated agreement). A lot of people know what this is, but very few do it, often because they don't even want to consider the idea that they won't get a deal done. You have to go there. If you don't play out the worst-case scenario in your mind, internalize it, and understand that if you can’t get a deal done the sun will still come up in the morning, you are in a much weaker place. Get comfortable with your walk-away position. 

These are two simple things that very few people do consistently. Following these rules will make you much more effective at getting to a deal that works well for both sides. 

Execution, Execution, Execution

Most sales teams don’t need more TAM to sell into, they need to get better at executing against the TAM they already have.

Most product teams don’t need more ideas, they need to get better at executing against the ideas they already have.

Measuring ROI In Enterprise Software

One of the main topics I talk to founders about is how to measure the ROI of their product and how to communicate that ROI to a prospect. This topic almost always comes up in sales conversations, and it’s important to be able to lead this conversation with clarity and authority.

I like to use a simple framework for how to think about a product's ROI, using three broad categories of measurement:

1/ Product usage and engagement. Registered users, monthly active users, transactions, data delivered, etc. Depending on the product, this can be more or less impactful. This is a useful way to think about ROI for a product that doesn't need to be used by a user (like an employee discount program or coaching software). This is not a very effective way to measure ROI for things like expense reporting or benefits management where users are required to use the product to accomplish something.

2/ User satisfaction. This is a bit of a step up over usage metrics in that it measures not just whether or not users use a product, but whether or not they like it. This can be an effective way to measure the ROI of an enablement tool where usage is not optional and financial gain is difficult to measure. NPS is a good measurement for this but I love the way Superhuman tracks this using this question: 1. How would you feel if you could no longer use Superhuman? A) Very disappointed B) Somewhat disappointed C) Not disappointed. There’s a great First Round article on this topic that’s worth reading.

3/ Revenue/Cost savings. This is of course the most impactful way to talk about ROI. It’s especially effective when a company is trying to create a category. In the early days of selling Zocdoc (an online appointment booking software for healthcare clinicians) revenue generated from the service was a crucial part of the ROI conversation. Most doctors didn't feel like they had to put their schedules online, so the only way they'd buy is if they were comfortable that they'd make money. While this was always important, it became less so over time. Online appointment booking became a standard. They had to do it. So other metrics and measurements became more important (e.g. does the staff like using it?).

Depending on the stage of category creation for your product as well as its competitive dominance, it’s important to understand where your product sits in the framework above. Some products need a hard financial ROI, others don’t.

The canonical example of the latter is Salesforce.com. A few years ago, I asked a Salesforce sales rep how they talk about ROI with their customers and he looked at me like I was crazy. The CRM category has been created and it’s now quite mature. Almost all companies of a certain size need a CRM. It’s sort of like calling Verizon and asking them about the ROI on your cell phone. At some point, you just need it. So Salesforce doesn't need to convince you that your sales teams will make more sales because you're using Salesforce, they just need to convince you that everyone uses it or uses something like it and that you need it too. They can validate their ROI by showing usage stats (the bottom of the stack). And if your team isn't using it, that's likely your own fault because you haven't done enough training or promotion to get employees to use it. And of course, they'll be happy to sell you a service that will do that for you.

When taking a product to market, it's important to recognize where your product sits on this stack. Are you selling something that will only be purchased if there’s a crystal clear ROI, or are you selling something that is required to keep the lights on?

___________________________________

Footnote: If you’re interested in learning more about category creation, I highly recommend the book Play Bigger by Al Ramadan.

Footnote 2: Generally, when talking about ROI you have the buyer and not the user in mind. However, it’s important to understand how both are thinking about assessing the ROI of your product.

Footnote 3: Eventually, all ROIs come down to dollars and cents. As an example, user satisfaction might lead to better employee retention which saves your customer money. But don’t go there if you don’t have to. ROIs generally have lots of assumptions that are easy to disagree on and challenge. Striving to show a financial ROI when it’s not needed can complicate/undermine the story you’re trying to tell.

Put A Stake In The Ground

When you start a new venture — a company, a team, a job, a product, a project — setting goals around its success can be stressful. You don't know how fast things will move and how successful you'll be.

Further, lots of people are afraid of being held accountable, much less being held accountable for something that isn't yet understood.

So there's a temptation to just get started without setting goals and see how things go.

For example, I've seen many startups not set sales goals in the early days because they feel like they don't have enough information.

This is a bad idea.

Setting a goal gets you and your team rallied around a target. If you meet or exceed the target, the team will feel great, and you can celebrate. If you miss, you can surface learnings and insights relative to the goal you set.

If you're hitting or exceeding your goals, surfacing learnings is less important. If you're missing, learning is crucial. A learning that isn't connected to a goal is much less powerful and much less interesting than one that is. This will also create the habit of being held accountable and reporting on failure as much as you report on success.

Put a stake in the ground. Set a goal. If you hit it, great. If you miss it, you'll feel a great deal of pressure to surface high-quality learnings that will get you closer next time.