The success of any company is determined by the ability of its leader to predictably and profitably grow revenue.

As CFO of a large public media company, venture capitalist, and investor and advisor, I've spent 30 years obsessed with projections. Many factors go into those numbers—from product to people to markets—but I'd argue that there is nothing more important for a company leader to be obsessed with than projections. When I was a CFO, I quite literally walked around with a printout of my up-to-date projections at all times. I don't ever remember taking that paper out of my pocket, because the numbers were imprinted on my brain.

I wouldn't say that every successful CEO I've been around has been equally fascinated with the numbers. But the big-picture, sell-the-dream types always had someone else obsessed with numbers at their side. 

Before digging more into the core questions you have to ask yourself about your model, let’s clarify the difference between two sets of numbers: the ones you actually use to manage and grow your business and the projections used for a variety of external audiences.

In most instances, when setting expectations with external audiences, it's a good idea to sandbag so things can go wrong and you can still make the numbers. I've been around plenty of CEOs who chronically over-promise and under-deliver, and it inevitably leads to a death spiral. They have to inflate the future more and more to make up for each successive miss until all connection to reality is lost. In these situations, as an investor or advisor, I find myself asking the question, "Where is the bottom, and what can I actually believe?" If the CEO fights me because they are too deep into their own denial, I know they are sunk.

The exception to the sandbag principal is when you are selling the company. This is a one-time exchange where you want to be plausibly optimistic about the numbers. When I was selling the Providence Journal Company to Belo for $2 billion in 1996, I was tasked with secretly meeting my counterpart to review our models and talk about price. Our Street estimates were appropriately sandbagged, so there was no way we would get our price based on them. My own model was a roll-up of 12 television stations, our newspaper, and our start-ups (like Television Food Network) ticked and tied to each operating budget. Higher than Street estimates, but still not high enough—I knew damn well that each of those operating leaders had sandbagged me. So I built and presented a set of numbers of my own creation that were the high end of what I believed our assets could generate. And those figures set in motion the deal that ultimately had Belo buying our company at two times our IPO price of just 90 days prior ($32.50 per share compared to $15).

The very last step in finishing the transaction was required by the buyer. Belo wanted to vet my model with our operating people to make sure they could deliver the numbers. These were folks who had no knowledge of the deal or my model. The most important employee had spent decades with the company and emphatically did not want to sell. But on a day’s notice, I had to drag them to NYC to meet the buyer at their banker's offices, trying to explain to them en route what the hell was going on.  

It was unpleasant and upsetting, but when push came to shove, they told the buyer that my numbers were certainly plausible, if not probable. I had pushed the envelope, but just enough to get our price without making anything up.

In fact, after the acquisition, we did make the numbers in my model. The deal was a huge loser for Belo (they recently sold the paper for $46 million), because they had stupidly decided to double down on a dying industry—not because my numbers were wrong.

Enough for the wind-up; let’s talk about the numbers you need to run your business effectively. Every business leader must constantly ask [redacted]. The goals are growth and profit. The question is how to achieve those goals with the highest degree of certainty.

The strength of a recurring revenue business is that revenue growth is generally (but not always) more predictable. Customers tend to be very sticky, and as a leader, you can isolate what causes them to leave (churn) and what it costs you to add new customers through sales and marketing strategies. That’s why the SaaS ecosystem has gone nuts with all kinds of metrics, as I have written elsewhere. Companies with limited competition also share the ability to predict future revenue and profit streams with a higher degree of certainty.

Whatever industry or business model your company falls into, the issue is how to think about the future of your company mathematically. Predictability is the key. There is a reason that Wall Street rewards companies that grow profits quarter after quarter after quarter. Short-term spikes mean little. This is a long-run game where sustainable revenue and profit growth is what matters.

There’s nothing magical in how I think about the levers in any company’s projections. I start from the bottom up: 

  • Focus on the product. What is it that you do for your customers? How are you planning to make it better? How much will that cost? What is the expected increase in revenue, all else being equal in terms of sales and marketing efficiency?
  • Think about existing customers. How good is your customer service/success function? How happy are your customers with the current product? Who is leaving and why? How much of an opportunity is there to upsell existing customers to grow revenue? Can investments be made in existing customer relationships to generate a positive ROI?
  • Consider new customers. Across the board, the most common problem I’ve seen with a CEO’s ability to predict the future is that they treat sales and marketing as a black box. On the last day or week of the month/quarter/year, they are still expecting a miracle, because that’s what their head of sales told them. If you break down the selling process (starting from the most basic marketing activities through lead generation and sales pipeline categories), there really shouldn’t be any surprises. Use history as your guide. And assume the worse while hoping for the best. You aren’t going to grow revenue predictably (assuming the same product set) without increasing both the resources you expend on the sales and marketing function and the efficiency with which those resources are expended.

After looking at things from the ground up, I generally take a top-down approach. I ask myself three critical questions: 

  • [redacted] Is my confidence level that we will hit those numbers 75% or better? If not, what do I have to do to increase my confidence? Obviously, smoking dope during this phase of self-reflection is a really, really bad idea.
  • [redacted] That is, will it generate the level of growth in enterprise value that my investors and I will be happy with after all the blood, sweat, and tears? If not, what do I have to do to increase the profitability ramp in the plan?
  • Finally, the X factor: people. [redacted] If I don’t, how long do I plan on treading water before making a change with each person I know to be sub-optimal? Does every group of people in the company clearly understand their role in achieving the projected outcome? Is there compensation tied to the performance you are depending on to make the numbers? Is there dead wood lingering in the belly of the beast that you are holding onto for some political, sentimental, or other external reason—anything that requires you to have the courage to do the right thing for the sake of the enterprise?

If you have done all the above, you will now have a financial articulation of your future to use as the roadmap for everything else. But alas, no plan is perfect. The world has this strange way of continually changing. So as I walk around with the model in my head—whether as CFO of a huge company or advisor to the CEO of a search fund—I am consistently asking questions that help me refine my view of my current model: testing, prodding, listening, talking to customers and low-level employees, and digging into product development. I constantly want to make my model better. Solidify the weak spots. Get better visibility on the top line. Figure out how we might increase profit growth in a sustained way. 

These are all the typical functions of being a leader. I just happen to think about them in the framework of my internalized model. Maybe I am unique in the way my brain is wired, but I don’t think so. I have found over decades that the best way to see the implications of every conversation and action is to run it through my mental model for the business. It leads me, at least, to a clarity of purpose that I have noticed in all the successful CEOs I have been around.

Of course, there are moments of temporary insanity in the capital markets when no number seems to matter. But if that is what you are banking on, I suggest you give up your CEO job, take all your worldly assets, fly to Vegas, and put it all on red. If you win, repeat. 

If you really want to grow your company, then it's time to get obsessed, day in and day out, with the numbers.