No plan survives contact with a customer, but you gotta have a plan!
Before we can go much further, we are going to have to answer some deep questions about raising finance for a business. Questions like:
- What is my business worth?
- What price should I sell shares at?
- Can I repay a loan, and if so, how fast?
- What will my business look like in 1,2, or 3 years’ time?
- What will this be worth to sell in future?
- What will the founders, investors and lenders earn from participating in this?
- What might change?
And, to answer those we need two things:
- A set of reliable accounts for the history of the company in standard (and IFRS compliant)?
- A financial model of the future?
Very few companies, and even fewer start-up and early growth companies really know how to make financial forecasts that help them manage their business, drive growth, and raise finance. The whole area of financial forecasting is probably two or three years of explanation, training, and exercises for a moderately competent accountant. But what we can hope to get over in this short article are some of the key features that any CEO or COO should ask for.
When it comes to financial modelling, most people only produce a simple profit and loss account on a cash basis. They work out what they will spend and what cash they will get in from sales each month and apply some simple growth assumptions to that model. I would urge anybody who is thinking of producing a financial model to work with somebody who has some accounting experience and who can prepare them a proper three-step model covering profit and loss, balance sheet, and cash flow. Taken together, all three elements give you a much better view of the way in which your business is likely to work. They also provide a much better basis for estimating the value of your company and the cost of investment.
A tidy model, with well thought out assumptions that are linked to solid evidence from independent sources, and which indicates the major drivers of value in your business is a good stepping stone towards a dialogue with the potential lender or investor. Wherever possible, the assumptions that you use should be external and activity driven, rather than just a simple “next month we will sell 20% more than last month”. Any investor or lender will have access to phenomenal large datasets of financial information from competitors in your sector, and they will very quickly identify where your model is not realistic unless you have done your homework properly.
The questions I always ask about a financial model are
- Is it entirely driven by external factors?
- Is it driven by activity levels?
- Does it identify every step in the sales to cash chain with appropriate and controllable KPIs?
- Are all assumptions clearly identified and reasonable against solid information?
- What is the learning curve factor that you are using?
- Are you 100% sure it has no hidden assumptions, constants or factors?
- Are all ‘links’ and factors identified and shown clearly?
- Is is capable of accepting any reasonable change to any assumption on a ‘Scenario Basis’?
- Does it record all changes, so it is always possible to roll back to earlier versions safely?
- Can it provide output in the same format as financial accounts to allow instant real time comparison?
- Can it also output critical KPIs, and value indicators (margin percentages, EBITDA, etc)
- Does the model integrate profit and loss, cash flow, and balance sheet correctly, after taking into account all known timing and taxation factors?
- How does it compare to the reality of competitor companies in your sector?
Most of the time I’m very disappointed by the answers that I get. A great deal of the information that people attempt to use to manage their companies is placed into Excel spreadsheets, literally my least favourite way of producing a financial forecast, which are then randomly amended over a period of time. This is very sad, considering so many excellent forecasting tools exist that can be integrated with financial systems. Xero, Iris, Sage, Dynamics and many others have such forecasting systems.
Very few forecasts I’m shown past the first three tests. There is a habit of “starting with the expenses” when producing a forecast. This leads to consistent, but ultimately useless models which bake in current inefficiencies and bad working practices for the long-term future. Many in-house accountants are good at showing internal assumptions. Things like “each person uses five pencils a year”. But then they bake in and hide some of the bigger assumptions, things like productivity, efficiency, staff turnover rates, customer retention, and fixed overhead costs.
What are looking for in the model is something that shows that it is based on the outside world and clearly walks through from the total addressable market, through exposure, to conversion, to sale, to completion of an order, through retention, to the collection of cash.
And this really is the key: do not let your accountant build the model alone. The model belongs to the CEO, the CMO, the COO, and the rest of the board. The process of working through all the activities and their dependencies and working out the people and resources needed to deliver them against a range of scenarios is a critical business discipline. I really cannot stress enough how important those sessions are to the management team, and to how investors and lenders will judge the maturity and capability of the team. The process of building the model can build consensus, iron out problems, find strength and weakness in the business plan, and identify where things are not yet known but evidence is needed.
Consider a stand making tea at a fair. The first team modelling session might go like this:
CEO we want to make a profit of £100 for charity by selling tea, what will that take? The model from last year is on the screen and we made £70 – ok, CFO will update on the screen while we talk.
CMO well, the fair has 2,000 guests, and last year 500 people bought a cup of tea from us at price of £0.50 a cup. Of those, 300 also bought a biscuit at £0.25p. We spent £10 on signage and paid the local youth club £25 to walk around with an advert.
COO It was a hot year, last year. If you remember, the year before, it rained for half the day and I remember that year we lost money. Last year, as normal normally rented a portable gas-powered water boiler and we rented the cups, saucers, plates, and cutlery. We also rented a small tent and some tables and chairs. We had a new team, and they took an average of 2 minute to make a cup and collect the cash. I watched a sample of 100 people and they sat and drank the tea for an average of 10 minutes, occupying 1.2 chairs each on average. From 11am to 2pm all tables and chairs were busy, and the queue was around 15 people long. The CMO and I then noticed that about 20% of people gave up and left the queue without being served and went to buy a cold drink instead.
CFO yes, the last accountant showed that we made £70 last year, after spending £70 on milk, sugar, tea and biscuits. We also spent £50 to rent a stand in the middle of the ground, the £35 on promotion mentioned by the CMO and £100 on renting the equipment and gas to heat the water. The problem is that previously any left-over stock of tea bags, milk and biscuits was carried forwards, and it has spoiled, so we lost £40 on that, meaning we only really made £30 last year. We need better stock control, at least.
CEO, ok, so we need to do 333% better this year. How?
CMO I would like to see if we can get a taller sign and pay the extra £10 to have a site nearer the busier attractions. I think I can reduce people leaving the queue if the COO and I can work out how to have the queue look shorter and move faster. Cold drinks are always popular, and we could sell those, but I don’t know what that means in terms of costs? I walked around the fair and saw that small cakes are super popular and also appear to have a higher profit margin than biscuits, maybe we can try them? I also think we should swap to paper cups and plates. People can then walk off with them, with our advert on the side of the cup. I will discuss that with the COO.
COO I can help with that. I also think we could rent a larger tent and two boilers. Yes, that is expensive, but I think it will work if we can get up to 1000 customers through, I’m concerned that the weather that month is 30% likely to be rain. A wash out would cause a loss, and we can’t afford that. And the tent can be shade on hot days. I’ve looked at seating and think we can get 40% more people through if we have smaller seats, some benches and some standing tables. I approve of paper cups. I think we can cut the time to serve by 50% and reduce the clean-up time and equipment rental. If we offer both sorts of cup, we can see what effect that has this year and fine tune next year. I can cut the waste if we pay that youth club to run to the shop at 12 and 2 and only buy what we need based on the weather, the customers and what we learn. I will check with the CFO on prices and see if that works.
CEO ok, that’s a good start, we have some scenarios to model: rain, sun, normal. We have some things to test with paper cups and standing tables. We need to find out some prices and costs. We might need another volunteer to serve. I will try to recruit two more volunteers with tea and cake experience. I wonder if we can increase our prices a little – can the CMO get local prices and come up with some estimates of price elasticity? COO, is there a better way to collect money and reduce the time taken to serve? Ok, CFO, can you get that model off the screen, update it and model those new bits and share them around with your comments, then we will convene next week with actions and see if we can tune this even better? Let’s see if that £100 target is possible!– made up meeting
Does that sound like your modelling sessions? Just remember: teams that model together, learn together and profit together.
The best teams are also really good at understanding that early-stage companies change everything as they grow. The learn. The market changes. New technologies can emerge. There are a number of non-linear interactions as companies go from start-up, through early stage, to growth, and to maturity. At a simple level, it is likely that the CEO will go from being the lead salesman, to finding 60% of their time is taken up with team and stakeholder management over the first three years. This has a dramatic effect on the sales profile of the company. On the other hand, the general efficiencies of teams tend to fall as they get larger and as levels of management are introduced. However, it may be that research and development gets more effective with slightly larger team sizes. It is these sorts of things that are important to model, because it is these sorts of things that the Board of Directors should be keeping in mind. Not the number of pencils that are used.
The idea of a learning curve is very poorly understood by most start-up executives. Generally, one has to have managed a substantially larger company, or significant project over a long period of time, to grasp it and to be able to express it in simple terms. At its very simplest, an assumption that every time the team goes through a learning cycle it will become 20% more efficient is quite a good one. However not all learning curves are positive: SaaS companies are all well aware that next year’s sales are likely to grow only 85% as fast as this year’s sales grew.
It is extraordinarily rare for any models produced by early-stage companies to take into account what can be learned from competitors and market sector reports. Yet this information is freely available and will certainly be held by any investor. If you have a model that shows you will be hugely profitable because you spend 6% on marketing and 5% on research and development, but your competitors have a market average of 30% marketing and 20% research and development then you will quickly have a serious credibility problem.
Fundamentally, most financial forecasts and optimistic hockey stick shaped revenue curve. I would say 95% of models I’m shown fall apart under the 13 tests above. Please don’t be one of those.
Some (almost useful, but they will insist on using Excel) resources:
If the CEO can get to a position where they have a forecast model on one hand, the actual outcome of the business on the other hand, and each of these is up-to-date as of 5 PM the day before then the CEO and the board will be able to make informed decisions to help steer the business. If not, they won’t. Business is too complicated, and memory to fallible, for people to “wing it” for more than a week. So, modelling really matters, building the model together from reality matters, but what matters more than modelling is having the ability to check reality against the model.
All of this is just laying the ground for the hard discussions to come.