Why your financial model should forecast 3 not 5 years

Tl;dr: Build a 3-year financial model and not a 5 year one. It’s easier to build and defend. More years mean more uncertainty and more questions.

How many years should you be forecasting in your financial model when building a model for investors?

This is a pretty nerdy question, and whilst it might seem trivial, it’s actually a core one before you spend a moment building a forecasting model, let alone buying one.

There are three main reasons here:

• You have to fill in the model. If you only want to forecast 3 years, but it’s a five-year model, you can’t say “ignore the last two”
• All the years you present are fair game for investor questions
• It’s really hard to change the model to a different number of years

We’re going to go into the details of the number of years in a model in this blog now so you understand what you’re actually doing. Awesome!

What are your options?

There are really only 6 possible options for the number of years in a model as I see it. A few are only there for continuity purposes. So quickly, let’s go through how many years you might have in a financial model… and just to be a nerd I’ll chuck in one for big businesses like oil pipelines:

• 1 year:  If you have a one year plan, it’s super short term focused. It’s more likely an operational plan used to give targets to the team. You would never show an investor a one-year financial model as they will expert more and it doesn’t answer the questions they actually have
• 2 year:  2 years isn’t necessarily dumb, it is obviously better than a 1-year model, but what purpose does it solve? I would expect a 2-year financial model to basically be a more detailed plan to execute on. Maybe it helps the CTO and product manager map out their roadmap? On the downside, 2 years is really not normal as things are likely going to change in year two. Yes, you can forecast year 2 fairly strongly, but to what end? I’ve never seen a two-year model.
• 3 year: Ok, now we are talking. 3-year models are very normal. You will almost always see a 3 or 5-year forecasting model. 3 years makes sense as it is just at the limit of what you can foresee since you might reasonably expect nothing to change, or at least, it’s sort of how people think. I’ll explain more about this in a minute.
• 4 year:  I could expound on a 4-year model, but really, it’s just a random ass number ;). I mean, you could sell a 13-month plan as a SaaS company, but customers would just think “that’s f’n weird!” right! I mean there I could make some logic for 4 years, but it’s just not something you want to consider.
• 5 year:  5 years is a very normal number when you are a big boy, in M&A and the like. Why is actually pretty academic and is for reasons that simply don’t apply to a startup. You see when you work in banking you are valuing a company. You don’t give a shite about what the actual operational numbers are. So, the actual output of a financial forecast is a DCF. In that, you discount the cash flows and then chuck on something called a terminal value. You typically forecast the model for as long as makes sense, or in so much as the client is paying… I could go on, but it’s not useful to you as a startup. I just want to explain why 5 years is a thing in banking. But if we are to be real, a founder is going to forecast for only so long as they can 1/ be bothered (let’s be real!), 2/ has a line of sight and can reasonably think through, and finally, 3/ be able to defend to investors which is all that you really care about when it comes to getting cash in the bank. I have and occasionally do 5-year models, but it is for a reason. I’ll explain more later.
• 10 year:  No startup is ever going to do a 10-year model unless they are an ex-finance type that thinks they know a lot since they were a banker, but has forgotten they’re in a startup now. There is zero value in doing a detailed ten-year model. You should use the future fundraise model to do this in a far simpler model and with very different objectives. In banking, you do a ten-year model when 1/ the CFO wants one, or, 2/ the industry justifies it and it’s normal. You are typically talking about insurance and industrial businesses. Again, I’m not going into details.
• 20 year:  Even most bankers will never make, let alone see a 20-year model in their career. It’s only when you do random crap like oil pipeline businesses you possibly would go here. The only two that come to mind are a pipeline business a team was selling and this random windfarm model this team hired a company to make

Why I build models in 3 years

Before I started building my big ass fundraising models, I spent weeks considering things you likely will spend zero, nada, zilch time considering.

The 3 vs 5-year financial model thing was a big one for me.

When you build a model from scratch it’s f-all work to really add 24 additional columns, or another 24 rows if you are making cohorts. I mean it’s like ten more seconds every time the 36 vs 60-month variable comes into play. There’s a couple of esoteric places like in the tax and deprecation schedules which take more time, but most things are just dragging formulas across.

I could easily have made every model I either sell or (for the most part) give away for free as 5 years.

But, I didn’t. And I know you want to know why since you’re reading nerdy f’n stuff like this! Ha. Love you too!

There’s one BIG reason why I use three-year forecasts and it is based on the fact that you overestimate what you can achieve in the short term and underestimate what you can in the long term.

You overestimate what you can achieve in the short term and underestimate what you can in the long term

That’s just one huge mind feck if you know enough to think through the considerations.

All investors want to see big numbers from founders as their business model depends on it, but deep down, given experience, they know that those numbers rarely happen. At least in the short term.

They always get surprised by the big inflection points.

Those inflection points happen when you don’t expect them, and for most companies, years in the future, unless you are a massive exception.

Whenever a portfolio raises another round they always make a humble brag post on LinkedIn with “great to support”. They’ve got a moment to make out they are betting on the right horses in the race.

It’s only when total and blind luck comes their way that they nail that certainly down with a Medium post to communicate to the next big founder that they’d be lucky to have them too.

I could actually argue that models with too many forecasts are not desired by investors as they require too much cognitively. I mean, VCs are meant to be experts, but they can’t really think what is really possible for a company, what regulation might come to pass, what competitive dynamics might arise, etc etc.

You’re asking too much of them and who wants to be called out even if no one is asking?

If you want to make friends at a dinner party you just need to shut up and let people talk. If they misquote Shakespeare, don’t be a dick and correct them.

In a random way, too much information stresses out investors.

Let’s look at what a 3 vs 5 year model looks like for fun.

Model with five years

Here we have the enterprise SaaS model. There’s a bit of the KPI sheet.

The first five years take up quite a lot of a page. I kind of feel it becomes the focus of attention. That’s fine if you are later stage company since the monthly progression is probably a lot less important than the annual one. Investors are likely focused on the yearly numbers and then go into the months just to check there’s no funny business.

 

Model with three years

Here’s the SaaS financial model. The three years are less domineering (Ok, the formatting is different).

I argue that it’s the monthly progression that matters a lot more when you are early stage. Sure, I want to know what you are doing at year-end, but as an investor, I would totally be looking at what happens month to month to see if anything doesn’t make sense.

But enough of investors, let’s focus on my darling you.

Late-stage companies

Even if you hit series-C and are raising $100m, if you are honest, you have no idea what forecasts you will achieve before you attempt to IPO (only to get cock blocked by CORONA 2.0), what you do have is an insight into what you actually have to achieve. What you think doesn’t matter anymore as you are sitting on so many liquidation preferences that you just have to meet expectations or you’re going not to be an angel investor at the end of it. The forecasts are the expectations you have to meet.

You’d like those forecasts to be as short as possible since you’d rather no records of what you said you’d do…

Also for public companies, it’s worth knowing that there are these folks called equity research analysts (aka brokers) who write ‘broker notes’. These are sent out to institutional investors to advise them on what to invest in. The brokers always do three years of forecasts, or sometimes one forecast (for the current year) and two more years of estimates. They don’t do five years.

These brokers are seriously smart dudes.

Early-stage startups

You’re not reading this blog if you are at series-C as you are near the top of the Layer Cake and you have minions to do this shizzle for you.

Nope, you’re in the shite pile hoping to be even considered to be included in the batter of the first rung of the layer cake. No offense, just being real!

In my experience, 99% of founders don’t know what they want to exit for, let alone how much they plan on raising at their next round and how much they will raise.

For most of you poor peasants, you just have no idea but are super hopeful! #blessed

The truth is you might be able to guess what your revenue will be at the end of the month, but don’t really have line of sight of 6 months. The further out you have to guess, the less certain you can be.

Let’s talk about confidence over the next 5 years:

• 1 year: Pretty much anyone can set goals for this year, right? I’m not going to make a joke, let’s just move on for brevity.
• 2 year:  If you have any managerial and planning experience, then doing a one year plan should be pretty bla bla to you. But what happens the year after requires a bit of thought. I mean I could intuitively just pick a number for year one, but for year two I need to think more broadly about strategic efforts, hires I haven’t considered, product development, and who knows what else. I mean, it’s totally doable, but it’s not as intuitively easy to just grasp at straws and come up with a number you are confident in for year one as it is for year two.
• 3 year: The actual point of this new year is how you logically get to here in a structured manner. I think we can agree year one has to be done so we can take it for granted, year two would require some work, but now year three? Fark, definitely will require more ‘strategic’ type thinking. You can’t rely on the gut, you need to think expansively and you probably don’t have a simple paradigm to draw on. Pause here, reflect and consider if you agree with what I’m saying here intuitively?
• 4 year:  You see, if you agree that year three was a little tough, year 4 is another exponent of thinking…
• 5 year: And then 5 is another exponent. I mean you are confused and not super confident in year 4, and not Alexander is making you think another level of abstraction?

That’s exactly the point.

Every year is exponentially more complicated and uncertain.

Warren Buffet is a smart but fairly modest chappie. At his core, he believes that he is different from normal people in that he understands compounding more than anyone else in the world.

You can’t defend when you aren’t confident

When it comes to fundraising, confidence matters. You need to not only be able to defend what you say with investors in a logical fashion but believe it so that you can imbue confidence in investors, that you believe in what you are doing.

Read: Founder confidence. How you answer investors’ questions matter

If you can agree that each year you become less confident and you don’t have to show too many years or forecasts, then why go there?

My point is just don’t.

If an investor asks why you are doing 3, not 5 (rarely, if ever) you can just share my logic with them, though probably selectively!

Eh,  If you want a template answer ask in the comments and I could invent something that sounds smart in 20 seconds.

If you present numbers that are backed up with a logical model, no investor is going to call BS on you. Frankly, few people other than me actually enjoy looking at financials!

The earlier you are, the more they are investing in the team, market size, traction, etc, and with regard to your model, how logical your numbers are… not what the actual revenue is in year 3!

You and the investor know that the forecasts will be wrong anyway, they just want to know they are logical.

Read: Why you are making a financial model for your fundraise is wrong

Why I don’t build 5 year models (as standard)

To drive the point home, I don’t do 5- year models because I’m lazy or I don’t have the know-how. I purposefully choose to make 3-year models because they’re advantageous to you.

I went through a logical exercise and decided that 3-year models make more sense.

If you have made a five-year model and you send it to me, I’m not going to berate you or think less about you. I won’t think you’re a moron or made a critical error.

The issue is that you have to make 2 more years of forecasts. It’s a lot more effort to do and you then have to defend those additional numbers.

We talked about uncertainty before.

• Year 4 and 5 is where you are expected to start hitting big numbers and they can make you cringe a bit.
• I know you haven’t a strategic plan for more than maybe 3 years, so you’ll have to invent them for the additional two years
• Your costs are going to increase a lot as you scale up revenue, and so you’re going to have to think more about your hiring plan
• You’re probably going to realise that your metrics don’t make sense, so you’re going to have to fiddle a whole lot
• It’s basically just a lot more for you to think about.

Thinking is overrated!

If you do your model properly, it’s going to take a lot of time. By the time you get to year 4 you’re not really going to care as much and you’ll get lazy. That can result in you making forecasts that might actually make you look like a dumb ass.

When I do a 5-year model

I’m the largest seller of financial models in startup land on the internet. As far as I can tell, I also have the honor of selling the most expensive excel model you can buy on the internet.

That model is five years. It is also for a specific reason.

The model I sell for 5 years is an enterprise SaaS model. I considered making it more than 5 years but reverted to the mean that founders can’t and don’t want to think too far in the future with so much uncertainty.

So why the heck even go to 5 years for any model you might ask?

Logic and the nature of the sales aspect of the business model.

It took me forever to build this fricking enterprise SaaS model and I did it for a company I was the COO at, whilst fundraising (well, the ghetto version). I was involved in a lot of sales, so I really understood the sales process.

One client told me the fastest we would be implemented was 13 months.

13 months!

FARRK! That’s a long sales cycle.

So if we can agree that 3 years is a solid forecasting period, then what the heck happens to any contracts you think you might close in year 3? I mean it’s year 4 and a bit before any meaningful license revenue occurs, right?

Whilst consumer and SME stuff might be super hard to imagine (let alone forecast), it’s actually fairly possible for an enterprise company who is a few years old to be able to plot out their execution plan with reasonable detail.

You see you can map out the landscape and then a ‘land and expand’ plan. You aren’t dealing with an abstract number of millions of SMEs, more likely there are 200 companies you are competing for and you have a dirty plan on how you will abuse them for revenue if you get in their… good graces.

Constraints are very empowering as you can imply a lot with it.

The huge irony in an enterprise is you can actually get a lot more certain over time if you have a solid vision of the future. Whether it holds up is another discussion.

a 5- year model is the minimum that makes sense for an enterprise SaaS company, so that’s why I make it 5. I could make it longer, but do you really want to have to defend a longer timeline?

What is the advantage of a 5-year model? AKA is there any?

Frankly, as a nerd, I can’t think of one that is useful.

I know there are founders, ex-execs, investors, whatever, that think a model should be 5 years, so the logic you need a 5 year model is that someone might ask for one.

As a supplicant who doesn’t know what they are doing, not conforming to perceived norms may seem to be something akin to fighting the man. You can’t win.

The funny thing is that I started out in ‘high finance’ where my bosses didn’t teach me shite but I was expected to know enough to write material for the CEOs of banks. I mean WTF.

I didn’t know enough and didn’t have the time nor luxury to explore the idea of challenging anything as I was kept at the whetstone as an employee.

Funny things happen when you are a nerd that doesn’t give a shite.

You can challenge everything.

If I was an ignorant kid, I would totally deliver a 5 year model if I was asked.

I’m not though, and I’m going to teach you how to not be either.

You can challenge anything with anyone if you have sound logic

Knowledge is power.

I call people ignorant all the time. They don’t understand what I mean and take it as a colloquialism (Note: Irregardless got added as an accepted word by an American dictionary yesterday, but anyway). I mean they don’t know.

If you understand, you can challenge it. Anything.

There is almost no norm you can’t violate with investors if you know enough to communicate your logic.

If you think 5-year models don’t make sense and you are asked for one, explain why you do a 3-year model.

What do investors want or expect?

I don’t know, maybe 20% of all my friends are investors.

Everyone that manages money is smart, but smart is relative.

If you got a job at a top management consulting or IBD equivalent (I know that is a small %), you are just as intellectually capable as 95% of investors (albeit you haven’t allocated specialty learning time).

There is no training course to be an investor so most are going to regurgitate something they read somewhere rather than actually forming an opinion on what they want.

If (most) investors ask for a 5 year model, it’s not because they actually want nor need 5 years, it’s because they think that’s what they’re meant to ask for.

There are morons who ask for business plans (mainly angels, and particularly those in finance), but that’s because they don’t know better. You’re free to tell them you have a pitch deck. If they insist on a business plan, you just tell them that’s not what startups do and move on.

Most investors don’t know what they want, so you can tell them.

There are of course norms you can’t really argue with:

• You need a pitch deck
• You need some kind of financial model
• You have to pitch
• You have to answer questions
• You can’t do a Zoom call naked

But, what you give doesn’t and won’t be the same as everyone else.

So long as what you do or provide is strong, no investor is going to call BS on you.

• If your model is 4 years, then they might think it strange, but they don’t care
• If your pitch deck is 35 slides, but really easy to read, that’s fine
• If you start a pitch by getting them to sit next to you and watch your code and it makes sense, who cares? (Twilio did that)
• If you have no supporting material to question, then who cares? If you have ‘sliders’ to FAQs, that’s fine
• If you work in fintech and go to meetings in a suit, that’s fine. Read: How founders should dress when pitching a VC your startup

So who cares how many years your model is so long as it’s too short. I just recommend 3 as it’s easier to do and easier to defend.

Changing models is really hard!

I mentioned that changing a model is really hard at the start of the blog (not that you can remember by now, LOLS). I’ll explain why now to close this blog off.

Let’s say you are building a house and the builder asks you for specifics.

You tell him that you want a 3 bed two bath.

6 months later you start to see what the house looks like and you go, nah. I want a 5 bed one now. Add two more rooms.

The builder is rightly going to stare at you blankly and say “sure thing boss“, once the initial anger subsides. He realizes how much more time and cost it will be for him to have two rip down the walls, change the foundations and add on two more rooms. He finally remarks “It would have been a lot cheaper if we did that from the start.”

The difference in building a 3 or 5-year model isn’t all that much, but changing it is.

Hey Alex, can you just turn this into a 5 year model and send back, please? Thanks!

To follow on, just wanted to explain that turning a model into a different number of years is a big deal.

When you’re really naive, you just don’t know better.

I get a lot of hilarious emails from founders and investors who request things thinking I 1/ actually work for them and 2/ what they are asking for is really simple.

Because I’m a huge f’n nerd, I throw out my little nerd giggle (to myself) at just how bonkers what they are asking for is.

I get asked about 8 times a year to turn one of my big financial models from a 3 year one to a 5 year one.

Let’s say there are 22 sheets and around 1000 rows per sheet… Do you have any idea how much monkey work is involved in changing that?

I’m going to guess 4x 15 hour days on El Chapo quantities of stuff that looks like asbestos but is not as fun.

Point is… a lot of work. And that’s for me to do- it will take you forever!

Conclusion

Do a 3 year model. It’s not the end of the world if you don’t. Less than 3 years is too little and more than 3 becomes harder to make and defend.