Top-Down Market Sizing

What is top-down market sizing?

Top-down market sizing (determining market size from above, also called the top-down approach) describes a method in which the size of a market is derived from an overarching, usually published total figure. This figure is narrowed step by step, through shares, to the sub-market that is relevant to a company. The starting point is, for example, spending on a product category, industry revenues or the sum of the revenues of known providers. The same principle is used by anyone who estimates the cinema tickets sold in a city by scaling down the nationwide sales figures by the city's share of the population.

The method is the counterpart of bottom-up market sizing, which builds market size from customer numbers and prices. Both routes serve mainly to quantify the total addressable market (TAM, the total market) and the serviceable addressable market (SAM, the serviceable market). The serviceable obtainable market (SOM), that is, the part that can be won in a period, is by contrast derived from sales capacity. In the business plan, pitch deck and information memorandum, the calculation from above serves mainly to place the order of magnitude and as a cross-check. In the pitch deck, investors expect at least the starting figure with source and year. In commercial due diligence, the review of market, competition and customers before a contract is signed, they then examine the justification for each narrowing step.

Top-down approach: procedure, formula and variants

The calculation consists of a starting figure and a series of shares. The first step is choosing the starting figure: a published total that describes the same category as the company's own offering. Then each narrowing step cuts away what does not fit the offering: other regions, other industries, other size classes, other product categories.

Formula: market size = starting figure × share of region × share of industry × share of size class × further shares

In practice, three variants have emerged:

• Spend-based calculation: the starting point is annual spending on a category according to a market study or trade association. The result describes the market volume, that is, what customers actually spend today.
• Provider-based calculation: the revenues of the known competitors in the relevant segment are added up and supplemented by an estimate for smaller providers.
• Structure-based calculation: the starting point is a macroeconomic figure, such as the total material or IT spending of an industry, of which a share is assumed for the service in question.

The method is quick and needs no customer data of its own. In early phases and in markets that are only emerging, it is therefore often the only route available. Its informative value, however, depends entirely on whether the starting figure describes the same category as the company's own offering and whether the shares are substantiated.

Sources and their reliability

For the starting figure, official statistics, trade associations, market studies, annual reports of listed providers and published annual financial statements come into consideration. For studies, four details are to be checked: the delimitation of the category, the reference year, the currency and the survey method. A study that counts hardware and services in the software market delivers a starting figure that is too high for a pure software provider.

The provider-based calculation runs into disclosure limits in Germany. Small corporations (size classes under Section 267 of the German Commercial Code, HGB), to which many young GmbHs belong, need under Section 326 HGB to publish only the balance sheet and notes, not the income statement. Small and medium-sized corporations may, under Section 276 HGB, combine their revenue with other items into gross profit. The revenues of many competitors in the legal form of a GmbH (German limited liability company) are therefore not visible in the annual financial statement according to HGB and must be estimated using proxies such as headcount. Such estimates are to be labelled as such.

Error propagation and limits of informative value

The weakness of the method lies in the shares. Every share is an estimate, and because the shares are multiplied together, their errors propagate and compound. Suppose each of three shares can be off by up to 25% upwards or downwards. The result then ranges from 0.42 times (0.75 × 0.75 × 0.75) to 1.95 times (1.25 × 1.25 × 1.25) the value using the mid-point assumptions. A figure given to the nearest million then feigns a precision that the calculation cannot deliver.

In substance, the method has a second limit: the calculation from above says nothing about how many customers a company can actually reach, through which channel and at what price. It is therefore unsuitable as the sole basis of a revenue plan. Its strength is the cross-check. If the calculations from above and from below land in the same order of magnitude, they support each other. If they lie far apart, the difference shows which assumption needs to be examined.

Worked example: top-down market sizing for maintenance software

The basis is the same example market as for TAM, SAM and SOM: maintenance planning software for manufacturing plants with 50 to 1,000 employees in Germany, Austria and Switzerland. All figures, including the starting figure, are freely chosen; they do not describe an actual market.

• Starting figure, freely set here and not taken from a real survey: worldwide spending on maintenance software across all industries and size classes of EUR 7.5 billion per year
• Share of Germany, Austria and Switzerland: 8%, so EUR 7.5 billion × 0.08 = EUR 600 million
• Share of manufacturing plants, excluding facility management, energy and transport: 40%, so EUR 600 million × 0.4 = EUR 240 million
• Share of the size class of 50 to 1,000 employees: 45%, so EUR 240 million × 0.45 = EUR 108 million
• Lower bound, if each of the three shares is in fact 25% lower than assumed: EUR 108 million × 0.75 to the power of 3 = around EUR 46 million
• Upper bound, if each share is in fact 25% higher: EUR 108 million × 1.25 to the power of 3 = around EUR 211 million

The reconciliation with the calculation from below presupposes the same delimitation. Calculated from below, the example market in this region and size class comprises 18,000 plants at an average annual value achieved of EUR 12,000, that is, 18,000 × EUR 12,000 = EUR 216 million. The ratio of the two values is EUR 108 million to EUR 216 million, that is, 50%. In marketing theory this metric is called the degree of saturation.

The difference is not a contradiction but a statement about the market: the calculation from above measures today's spending, the one from below the potential. The two results fit together if about half of the plants already use comparable software at comparable prices and the other half work with spreadsheets or paper. This can be checked against the company's own sales conversations. If only a small share of the plants use software there, either the starting figure is delimited too broadly, or the existing users pay considerably higher prices than assumed. In that case the price assumption in the calculation from below would also have to be reviewed.

Distinctions: calculation from above, TAM, market volume and market analysis

• TAM, SAM and SOM: the results of market sizing, not its method. From above, TAM and SAM can be estimated only approximately because the calculation usually measures today's spending; with the same delimitation it tends to give a lower bound of the potential. The serviceable obtainable market can sensibly be derived only from sales capacity.
• Market volume: the sum of today's spending by all customers in a market, regardless of which provider. A spend-based calculation from above delivers market volume, not market potential.
• Market analysis: covers, besides market size, also segments, competition, drivers and barriers to entry. The market analysis uses the calculation from above as one of several building blocks.

Common mistakes

• The one-percent argument. The statement that even a small share of a large market yields high revenue is no substitute for a derivation and does not explain where a single customer comes from.
• Wrong category. A starting figure that includes hardware, consulting or adjacent product categories overstates the market before the first share is applied.
• Outdated starting figure. Studies without a reference year or from surveys long past are not accepted in review. Extrapolations need a disclosed growth rate.
• Double counting when adding up. If sub-markets from different studies are added, their delimitations often overlap, and the same revenue is counted twice.
• False precision. A result from several estimated shares belongs rounded or stated as a range, not to the decimal place.
• No cross-check. Without a calculation from below, it remains open whether the order of magnitude is right, and the figure cannot be substantiated in review.

Frequently asked questions

What is the difference between top-down and bottom-up market sizing?

The calculation from above starts from a published total and narrows it through shares. The calculation from below builds the market from customer numbers and price per segment and is therefore easier to verify. From above one usually obtains the market volume, that is, today's spending, from below the potential. Investors see the calculation from below as the load-bearing derivation and the one from above as a cross-check.

When is the calculation from above useful?

To place the order of magnitude, as a cross-check on the calculation from below and in markets for which there are not yet countable customers or prices. It is not suitable as the sole basis of a revenue plan.

Which sources are suitable for the starting figure?

Official statistics and trade association data, because their methodology is disclosed, plus annual reports and published annual financial statements of providers. Paid studies are usable if delimitation, year and method are known and the figure can be evidenced with a reference.

What does it mean if the calculations from above and from below lie far apart?

That at least one assumption is wrong or the two calculations measure different things. First it must be clarified whether delimitation and measure match, that is, spending or potential; only after that does a remaining difference point to a faulty assumption.

How precise must the result be?

As precise as the shares allow. With several estimated shares, a rounded figure or a range is more informative than a single, apparently exact figure; what counts is the order of magnitude.

The Roemer Capital view

In mandates, Roemer Capital uses the calculation from above as a control calculation, never as a load-bearing figure. In the financial model both routes sit on a common sheet, with the same delimitation and with the source of each starting figure.

The difference between the two results is informative for Roemer Capital. If it can be explained by the difference between today's spending and potential, it becomes an argument for the equity story, that is, for the growth case made to investors. The statement is then: a large part of the market has not yet been tapped by any provider. If it cannot be explained, Roemer Capital clarifies the cause before the documents are sent out, because investors pick up on exactly this point in review.