Case: Cost structures and profitability of expert companies
In this review, we examine the cost structures of Nasdaq Helsinki’s expert companies. The purpose is not to rank different companies, but to examine the differences in the cost structure and open possibilities for examining them. The review includes Aallon Group, Digia, Etteplan, Fondia, Gofore, Innofactor, Loihde, Netum, Nixu, Siili Solutions, Sitowise, Solteq, Solwers, Talenom, Vincit and Witted. The group clearly focuses on the IT service sector, but it also includes expert companies from accounting firms to engineers and legal services. Other expert companies could also be found on Nasdaq Helsinki, but we limited the examination to medium-sized companies (excludes Tietoevry) and to those whose revenue mainly comprises services. The companies in the group have significant differences both in terms of the sector and business model, which means all comparisons should be taken with a grain of salt. However, there is a lot of similarities in the operations as well.
The success of expert companies always starts with the skills of the personnel around which services are created. Often these services are sold purely at hourly rates, but the company can also sell larger projects and price them on a value-added basis. Large fixed-price projects are rarely carried out today because of the risks involved. Often repeated services can be productized to make them more reproducible while enabling more extensive service processes. Especially the IT service sector also aims to create ‘generic’ tools or platforms on which customer-specific solutions can be built. In practice, all expert service companies try to create clear frameworks and processes for their services to deliver services efficiently and reduce dependence on individual persons. At the same time, the companies aim to improve their profitability and thus seek scalability that is characteristic for product companies.
Personnel costs account for a majority of costs
In practice, the largest cost item of expert companies is always personnel whose expertise and work contribution are the basis of the entire business. In the group we examine, the average share of personnel costs in all operating costs was 71% in 2021 (2020: 73%). However, there is a lot of dispersion, and especially Witted stands out from the group. The company uses exceptionally much subcontracting, which explains the very low share of direct personnel costs (31.4%). In the IT service sector, virtually all companies have different sized subcontracting networks, which typically transfers expenses to the materials and services category. In reality, the majority of a subcontractor's costs are probably personnel costs, but to the 'principals’ the cost is seen as purchases.
For the other companies in the group, the share of personnel costs in operating costs varied between 58-81%. Personnel costs are by far the largest cost item for ‘pure’ expert companies. This means that companies are subject to wage inflation, and staff productivity as well as utilization are critical factors for profitability.
All graphs included in the text are based on 2021 and 2020 figures reported by the companies to which we have not made any adjustments. Therefore, any exceptional expenses may distort figures and discrepancies in reporting their comparability. We did not consider depreciation and write-downs as operating costs in the calculation.
A high proportion of personnel costs may indicate that the costs are largely directed at persons performing the work. On the other hand, personnel costs can also arise form ‘unproductive’ organizational flab in middle management, so it is impossible to draw direct conclusions. In addition, many companies in the group also have product-based revenue and related costs. In principle, a high proportion of personnel costs reflects the tight cost structure and good focus of the service business but is not automatically the key to happiness.
Construction of expert companies’ profitability
As the product sold by expert service companies can be simplified as the expert's time and the biggest cost item is the salary paid to the expert (including overheads), their difference is a critical factor for profitability. We can calculate this ‘personnel margin’ by proportioning the companies’ revenue and personnel costs. The average for the group was 39% (2020: 38%), which means that experts on average generate revenue with a higher margin than this - but the entire personnel do not perform customer work. With this income, companies must be able to cover all other expenses and hopefully also generate return for shareholders.
We already discussed Witted’s anomaly, but other companies that stand out from the group are Talenom, Loihde, Netum and Gofore. For Talenom, a significant factor is the technology with which the company has been able to automate accountants' routines and thus significantly improve the efficiency of personnel processes. Loihde's business includes a significant amount of physical purchases in the Security Solutions business area which distorts figures from this aspect. Of the ‘purest’ service companies, Gofore that is performing excellently is joined by Netum in terms of this indicator and the ‘personnel margin’ of both rounds to an excellent 44% level. We believe that a margin level of about 35% can be considered good for expert services, so the average profitability of the group is very good with this indicator. We stress again that the figures are not comparable because different IT service companies use various degrees of subcontracting and some have significant product-based revenue. In addition, e.g., Etteplan has bigger purchases than the group on average.
Of course, the ’personnel margin’ is not a constant and can be increased, e.g., by raising prices (more revenue per hour worked), improving staff productivity or utilization rates (more billable hours) or by lowering average personnel costs. The latter is a theoretical option at a time when the competition for experts is tight in general. Optimization in this area alone is not always essential, but it is generally a very important area. The margin level also places a theoretical maximum on the profitability of the service company, as the company cannot avoid it.
Subcontracting included in the ‘gross margin’ approach
Another way to look at the ‘gross margin’ of expert companies is to calculate the margin after materials and services and personnel costs. The basis for this indicator is that companies often subcontract part of their work through their subcontracting network, which means materials and services are directly comparable to personnel costs. This is naturally not always the case, but especially in the IT service sector subcontracting is commonplace and other procurement is usually limited. This indicator will, to an extent, put different business models on ‘equal footing’ making comparison more meaningful.
The indicator also shows the scalability potential of earnings that the company could achieve. Thus the figure is also a good yardstick for pricing power and getting personnel at reasonable costs, because it quite directly tells how many percent is achieved between customer price and direct labor costs. This indicator is by no means a constant and the same principles largely applies to improving it as to improving the margin after personnel costs discussed above.
The median for the group was 25% (2020: 25%), which means that experts on average generate revenue with a higher margin than this also considering subcontracting. Subcontracting margins are typically significantly lower than for own personnel, but using subcontracting offers more flexibility in demand fluctuations (variable vs. fixed costs). In addition, some companies with product-based revenue or who capitalize personnel costs related to product development costs easily stand out as better in the category than pure service companies.
The company that clearly stands out in the group is Talenom, whose business we regard as a hybrid between a pure service organization and a software company. Talenom clearly stands out in the group in terms of its scaling potential, but it should be noted that the company's profitability is also supported by significant capitalization. According to our estimates, Solteq's high figure is based on the same factors as Talenom's. Among purer expert companies Aallon, Fondia, Gofore, Netum and Sitowise stand out and reach 26-28% levels. Fondia has very little external procurement, its average added value of services is high and they are partly productized. We also believe Aallon and Sitowise sell more solutions, so customers are billed more based on the generated added value. Netum and Gofore stand out, especially when compared to other IT service companies, which may indicate good billable utilization, but also, in part, the pricing power of the companies and the ability to secure labor at reasonable costs. Both companies also operate heavily in the public sector and have had low turnover.
Witted, Loihde and Nixu in turn operate at lower levels. The reasons for Witted and Loihde were already discussed. The profitability potential of Loihde’s physical protection (Inderes’ estimate: good 50% of revenue) is structurally much lower than for the others. In recent years, Nixu has struggled with high turnover and weaker than expected demand.
Other expenses indicate general operational efficiency
Direct personnel costs are the most important part of an expert organization's cost structure, but they are still only one part. Another significant and interesting cost item is other operating expenses, which typically include, e.g., office-related expenses, IT costs, sales and marketing expenses (unless specified), and a wide range of other cost items from travel to various purchases. To simplify other expenses create operating conditions for the organization (‘infrastructure’), and often these costs are relatively fixed in nature. This means that other expenses often scale as revenue increases, at least partly, but on the other hand they are not that flexible if revenue falls.
We note that we do not consider the differences between IFRS and FAS accounting standards in the treatment of rental expenses. Under IFRS, long-term lease expenses are mainly carried through depreciation and partly through financing costs, whereas under FAS they are recognized as other operating expenses. The use of FAS typically increases other operating expenses in service companies by a couple of percentage points, but, e.g., Loihde has communicated that the difference is 3-4 percentage points for EBITDA. In general, other expenses should be related to the company's revenue, as we have done below.
The median for the group is 11.6% with 2021 figures (2020: 11.7%), but there is a lot of variation. Once again, Witted stands out, whose other expenses relative to revenue were only 6.9% last year. Witted's subcontracting network probably requires slightly less other operating expenses than companies that rely more clearly on their own personnel. On the other hand, Witted was listed only in 2022, which is likely to raise the cost burden somewhat. There is room as the company has continued to grow quickly. Otherwise, Digia (8.6%) and Sili (8.8%) have the most efficient cost structures, in addition to which Etteplan and Talenom are below the 10% threshold. The 10% level is in general seen as a type of yardstick for an efficient organization, especially in the IT service sector, and we feel that it is a good target for the service sector. The figures are not fully comparable, but they provide guidelines.
Companies that perform well in the category (excl. Witted) are clearly larger measured by revenue than the average companies in the group. Size usually supports the scalability of other expenses, as certain fixed costs are automatically generated especially for listed companies. This view is reinforced by the fact that the highest figures are found among the companies with the lowest revenue in the group: Fondia (24%) and Aallon (16%). Both of these are also applying the FAS standard. However, Netum reaches the median level despite its low revenue and, e.g., Vincit fares negatively in this comparison despite its reasonable size class. In general, however, the scalability of fixed costs provides another good basis for targeting further growth and supports acquisitions. Fondia has the largest share of other expenses (as much as 24%), which is a very high level for an expert organization. We believe the company's infrastructure has been built to carry significantly higher revenue than currently, which is negatively reflected in the company's profitability for the time being.
Summary of profitability comparison
In this text, we have simplified the cost structures for expert services to the key factors, i.e. personnel costs and other expenses. To consider subcontracting, we also included materials and services, which enabled us to get the companies on a more equal footing. We believe that personnel/gross margin reflects the company's pricing power and ability to find skilled personnel at a reasonable price. The company's attractiveness to employees is very important when the company sells employee expertise. We feel the ratio of other expenses shows how trimmed the cost structure currently is relative to the size of the company's business. Here, smaller companies are on average weaker which also partially indicates scalability of profitability.
In the following graph we present the margin level after personnel costs and other expenses (light tone) and the companies’ reported EBITDA margin (blue). This again leaves depreciation and impairment outside the examination. The median of the group’s ‘combined ratio’ is 28% and the median EBITDA margin is 13%. However, the range is huge, which is influenced not only by the different profitability of companies, but also differences in the cost structure. Witted operates with a different business model, which means we could have excluded the company from the comparison. On the other hand, the company shows that there are different models and highlights how individual figures can be misleading.
We estimate that the differences in the figures are, in general, largest in companies where the importance of subcontracting is high. This cost item is usually classified as ‘materials and services’ which in some cases is very small (e.g. Talenom and Fondia) and in others very large (e.g. for Witted over 60% of revenue goes to external services). This again underlines that individual parameters cannot be compared without knowing exactly the differences between the business models. It is interesting to examine various factors, but the whole is decisive.
In general, the correlation between figures is high, which is natural when both usually represent a share of the entire cost structure. In addition, both parameters help investors understand the dynamics of service companies and potential profitability, provided the differences are considered. The logic of the cost structure of expert companies is very simple in that everything starts with the difference between billing and personnel costs. With this margin, service companies must be able to cover all other expenses in order to also have profits to distribute to shareholders. The analysis can easily be deepened by specifying in more detail which costs are fixed and which variable.