Many financial models and therefore forecasts follow an approach of simplified aggregated view, percentage improvements, or disconnected assumptions, making them high effort and low impact tools.
Alternatively, financial model, can become a primary decision making tool, that provides not only accurate assessment into the future, but allows to evaluate business on the operational level and quantify risks, all on a continuous basis.
Following the below descried approach will help any organization achieve these objectives.
Adhering to business processes decompose financial statements items into its most crucial elements, which behave differently, even if they are recorded as one category, or relate differently to other levels in financial statements. This might include breaking down sales into clients or product groups, clients into client categories, costs into subgroups, payroll costs into teams and compensation components and so on.
It's important to follow the business logic. For a concentrated B2B sales company, it would be more practical to provide insights into individual clients performance and then product categories, than just product categories without knowing who is or will actually be generating sales. Alternatively, for a B2C business, customers will be highly variable, but depending on the market, they will be making similar product choices, therefore location and product breakdown would be the better option.
Supplement financial items with explanatory operational data such as volume and prices at a minimum. Also include important contractual information, like payment terms, turnover discounts, staff bonuses, ect.
Its crucial to reconcile operational data to financials. Its best to use only complete and accurate data. Depending on the data readiness, it would be better to use more aggregated data or switch to a different source which reconciles better than use highly detailed but partial or misaligned information.
Provide monthly historical actuals as far in the past as it makes sense, to establish relationships among items and highlight trends.
One of the relationships example can be, how much resources it takes to provide services, products or certain desired outcomes.
Trends, help highlight overtime development of each item in particular, or relationships between items.
In some cases, relationships can be established on operating level, not financial, which would provide better explanatory power to the model and allow to design more transparent improvement initiatives. For example, one sales or customer success consultant can handle X amount of inquiries per day, is a better ratio than it costs $X to close the ticket.
In a separate section or in an analytical section pull all relevant items together, as it was one complete, yet separate, process.
Combining all the above steps in one, for example can look like that. One customer has constituted significant percentage of the overall sales, has been granted a special price, pays in foreign currency, shows seasonality in its purchases. Fulfilling its orders is going to require certain volume of inputs and labour at certain unit cost. With all information avaliable, this client performance, at all P&L and balance sheet levels may be assessed separately. Following this approach, the model can be broken down to operational performance of each individual client (with the minor clients as one item) and reaggregated back to reconstruct accurate high level financial picture.
Most companies will have revenue analytics by customers, or products readily avaliable, but only aggregated costs of sales. It might be more practical, for a company, to use aggregated costs, as the delivery process might be running continuously and be allocated to certain customers at the end, or might be continously switching between different needs. Following the above example, adding together all sales of a certain category, will define output levels and when compared with cost of sales' individual categories, will highlight standard resources consumption levels such as labour, materials, services, ect, as per actuals, at company specific terms. Extanding these relations into the future, the forecast can become specific and accurate, eventhough it quantifies relations at different levels. It's also readily avaliable for benchmarking vs the market.
Moving to forecast, automate calculation following the business logic wherever possible. What comes first in business becomes qualitative level input, decided by the company leadership, while all other lines can be estimated based on previously developed relationships with this category.
For example, for a streaming service, it can take 100GB and $7 per GB of bandwidth cost to provide service to one client. Therefore, next month expense becomes a derivative of the number of subscriptions, average users activity and price per GB.
In a different case, new clients, can be described as a result of an online promotion spend and historical conversion ratio. In the absence of supporting data, simply historically observed monthly trends of new customers, can be aa first level input. In many cases simpler is better.
Its important to maintain cause-effect logic at any point and neither over automate or over manualise the first level inputs. In certain cases it’s advisable to aggregate information to a higher level to find stable ratios and trends. In other cases its better to dive deeper to define unique drivers.
Maintain ability to adjust inputs, even if they are the result of relations and automatic calculations.
Depending on the ability and a will to influence them, some variables can become objectives for improvement in the future, or serve as a sensitivity analysis levers. Its advisable to differentiate them from other automatic computational cells, or highlight as drivers, so they can be easily identified and changed.
In general, in a normally functioning markets, availability and prices of resources should mostly follow historical trends, unless there are developments that would require further evaluation and adjustments.
Few example from recent history. Pandemic lockdowns created all sorts of unexpected shocks to either customer behaviour, availability of inputs, prices, financing costs, ect. Assessing impact of macro and industry developments, can be a useful element of accurate forecasting, otherwise it can be assumed most items will follow trends and therefore its better to simplify the forecast.
Following the trends, maintaining relations and adding high confidence elements create a base line forecast of the business “as is”.
This most likely will be an final outcome after a year of work, for many reasons.
Clients’ behaviour and preferences are hard to be influenced
Trends persist as they are a vector of multiple forces
Every organisation developed certain knowhow and follows certain processes which are subject to inertia
Everyday, everybody is already busy doing their jobs, so without new approach the overall outcome will most likely persist unchanged
To clarify, this doesn’t mean stagnation or flat P&L. There are successful businesses that simply keep adding clients every month/year with the current strategy. This will most likely continue.
There are also businesses that that keep loosing revenue. This also will most likely continue.
Its also important to follow nominal scales, or simply count units. Many new businesses experiencing high growth from low base on single market, assume they will follow similar growth every next year, for example 100% as per the last year. Eventhough there might be a network effect at work, it’s also equally possible it will just be 1000 new customers every year, therefore the growth percentage will keep decreasing from 100% to 50%, 33%, 25% and so on, following lineral growth path. Suprisingly, if things go the worng way, they follow the logaritmic scale.
At this stage the forecast can be optimised for new initiatives.
This becomes management assessment stage. It involves, evaluation, generating ideas, defining objectives, assessing feasibility, concluding on short and long term objectives.
It’s also the most difficult and biased part of the forecasting process. Many assessments and amendments, can represent wishful thinking. Clients storming the gates, efficiency improving, costs decreasing, projects delivered with no challenges ect.
There are multiple reasons for making wrong assumptions and also multiple ways to eliminate biases, but it’s beyond scope of this article.
Calculate the risk levels of the plan
Taking the above mentioned reality into consideration, in some cases it’s advisable to have an alternative scenario in place, assuming that very optimistic plans will simply not be delivered or delivered only partially.
One might say that if the plan is unrealistic, why not make a realistic one and follow it. Besides above mention reasons, there is also one paradox which actually makes overly optimistic plans not entirely wrong. Without a vision there will be no progress, without a believe in success there will be no effort, without an attempt there will be no result. Thus in some cases, or in some areas, the plan needs to be unrealistically visionary.
But on the other hand, it also needs to be realistic by incorporating assessment of its delivery difficulty, to ensure health and overall wellbeing of the company, in either case.
The best method to overcome this paradox, is to develop and compare both plans against eachother and define ‘accounting value at risk’. How much revenue, how much profits, how much cash seem to be attributable to improvement initiatives above the base line scenario. This would help to structure operations, implement contract terms ect that would support both outcomes at once.
When the final plan is agreed upon, company has successfully developed a budget. A vision how things should evolve overtime. Agreed to be delivered as such by owners, management and teams. Moved to an execution phase.
A forecast is always only from ‘the current day onwards’, therefore with changing circumstances, the last month's forecast becomes irrelevant and only a new forecast should exists. Rolling the forecasts based on the currently delivered actuals becomes the new target.
Some numbers can simply move to next period, or might never be delivered. Thats a big change. Therefore after a quick update, followed by calculating variations between sum of actuals and a new rolling forecast and comparing versus the budget is the best tracking method for overall business performance assessment overtime.
As a summary paragraph, following this approach significantly helps to eliminates the usual challenges in business planning and ensures forecasting accuracy.
Forecast is driven by few well defined inputs on operational level and therefore is quick to prepare and easy to evaluate.
There is no high level unexplained percentage changes, which are not attributable to any particular item, initiative, team or person.
There is a clearly defined vision that requires step by step execution.
There is clear assessment of execution risk.
There is a quantification of high effort and status quo paths, where with high confidence, the final outcome will most likely materialise.
There is a continuous assessment where the company is heading and how does it follow previously agreed upon plan.
How does it work in business reality based on few examples.
For a global shipping company, 1 hour assessment of key customer purchases, supported by historical figures, let finalise budget and assess all key performance risks.
For an online entertainment company, the baseline forecast had no adjustments, only discussion on seemingly unfeasible high growth projections, which was concluded to be accurate after reviewing trends in customer growth and their interaction with the platform.
For a real estate company, base line scenario which followed long term trends and challenged overly optimistic vision, become a spark for multiple corrective actions ahead of time, ultimately saving and streamlining the business.
SaaS company, which rejected to consider baseline scenario, followed its path to bankruptcy.
For a B2C company, forecast discussion led to multimillion savings in promotion activities, redesigning of go to market strategy folowed by three digit sales growth in the following years.
Any company can follow this approach. Depending on the system availability, it can be implemented in either ERP systems, FP&A platforms, online or offline spreadsheets or a mix of the above. It transparently presents the current performance and allows the company to accurately plan for the futuure.
Many CEOs and owners are led to believe that model is a model. There are two main types of financial models. Fundraising models and controlling-forecasting models. On surface they might work similar, but they are as different as SUV and two seater convertible. One used to drop off minors to kindergarten and the other, to take a loved one to a beach club.
Starting with Fundraising models. They are:
static
not easily updatable
focus on strategy
visualize markets and go to market strategies
estimate growth path
benchmark company vs market
estimate value created from growing the business
connect current track record with expected growth after funding
justify investment
list investment allocations
blend investment allocation into financial statements
present industry specyfic performance ratios
Fundrising models are a pitch to investors in their prefered language. Financial language.
Depending on the business stage, some founders when thinking about fundraising, believe they just need A model to be included in the data room. Draw up logarithmic revenue growth in one or few lines, don't think much about strategy how to achieve it, market constrains, ect. Just throw-add some marketing expenses. Breakdown salaries, rent, subscriptions/hosting. Draw up a simple P&L and keep increasing gross margin from the time product development is finished. Put few graphs from the model into a pitch desk, make it look nice, and go on a roadshow.
Even though, this strategy worked some time ago, when investors were following statistical investments methodologies, such as investing in multiple start-ups, expecting only a few reaching very high valuation to cover for all other unsuccessful investments. At that time, money costed close to 0%. Currently, with high quality bond's yielding between 4-5%, limited partners have an alternative to earn c.a. 30% return over 5y time with very low or no risk. Thus, to justify the investment in a new or existing business venture, the entire ecosystem needs to step up its return game. Therefore...
Funds started to focus more on the fundamentals of a business. A start-up is expected to present a vision supported by a believable financial pitch via its current P&L trends and reliable forecast. The later the round the more important it becomes.
Among many things, one that blocks many fundraising deals is the disconnect between performance till fundraising and the submitted projections. Sort of a V-shape P&L, which simply doesn't convince financial people.
There is no simple answer how the P&L should look like. It depends on the product, stage of development, addressable market, competition, final objective, operational ratios, and many, many, more. What is common, though, is the need of alignment between the vision, narrative and numbers. This doesn't happen over night in a spreadsheet. Understanding, how the success can look like, what are the constrains, where is the likely finish line, are all necessary beacons in running day to day operations. A financial model lighthouse. Overtime results of this work will support the vision with numbers.
Second important area is prudence in spending cash between rounds and execution excellence. Milestones, finished solutions, market achievements, all add believability to dilligent governance of the funds. Many times, engaged in daily challenges, this point is overlooked. One other way to express this point is look at it by switching from narrower to a wider perspective. An employee might have projects, a team might have projects, the company might have projects, but for investors the overall outcome is a project. Every single mentioned project has time and capital constrains. Bigger projects' problems start, when small projects underdeliver.
In the case the company is facing challenges in fundraising, the reason would most likely lay somewhere there. Although, it would take some actions, its still not a lost cause.
The multiples has normalized and stabilized. Whatever game the start-up is playing, meaning which valuation multiple would be the most relevant, it has to deliver it. That means, that after spending previous round, business needs to be worth more than it was before accordingly to new multiples.
A quick example. For one of the sectors, over last few years, multiples decreased by 50% from 8x revenue to 4x. A 10m revenue start-up, that after funding grew 20% per year for two years, reaching 14m revenue and meanwhile burning the whole round, suddenly its worth 24m less than previously (10m x 8 = 80m vs 14m x 4 = 56m; 80m - 56m = 24m). That felt like a rug pull for many. Up round is not justified by valuation. Down round is not an option for founders. Stalemate. All this is preventable, with the continuous financial management and guidance.
How to prepare, than?
Preparation need to start with internally developing transparent controlling-forecasting model, which will clearly and accurately show how cashflows are structured. Some of the numbers will end up in the fundraising model, making the same impression. Better to see them as soon as possible.
Through such a model, a company can continuously assess how it earns and spends money. Is it still on track with the vision. Not going into a lot of details, a reliable model like this, with all the operational and financial data, will provide continuous monitoring of the next fundraising timeframe and expected valuation at that time. Its important to remember, that at a funding moment, new valuation should be bigger than previous one, on a diluted basis.
If these two things don't align, most likely there are strategic or execution blind spots, on both the revenue and costs side, which the current team isn't fully aware of. Most businesses, would expect necessary adjustment to be on the costs side, but its also not uncommon, that small companies don't value their products correctly. Simply, undercharging them to a level, at which selling becomes difficult.
Arriving to a controlling-forecasting model features.
Controlling-forecasting models are monetary quantification of all company initiatives.
dynamic and easily updatable with actuals
utilize advanced modeling solutions
designed as a rolling forecast
operationally focused
monitor project delivery
tracks execution
contain a lot of data
present multiple angles for different departments
contain and track budget
explain variations
monitor growth trajectory need for valuation increase
With the support of such a model, company can identify issues, take adjustment actions, present few months of improvement on the P&L. Smooth the V-shape P&L, extend the runway, reduce time pressure, help communicate adjusted strategy and by making future projections more believable, in overall, ease the fundraising discussion.
Its tempting to say that the next fundraising starts, at the moment when the previous fundraising has been completed.
Using controlling-forecasting model, the company can stay on track with spending new capital injection and make the next fundraising round easy.
How it works in practice:
Multiple businesses, asked to help develop internal controlling-forecasting model acting on request by a funding partner
For multiple B2B start-ups, identified both pricing misalignments and costs leakages
For a multiple B2B start-ups, using controlling-forecasting models, identified deviation from necessary valuation growth, estimating challenging fundraising processes, around a year ahead of time
Received few 'best financial model ever seen' recommendation from global VCs
For a B2B FinTech start-up, through fundraising model justified a vision allowing for 32x revenue multiple fundraising round.