IIIOrganization & Project ManagementGuide
What Software Really Costs: Licenses, Implementation, Upkeep
Twenty-five euros per user per month sounds like a small decision. Add setup, training, upkeep, and the weeks with two systems in parallel, and the license is often only a third of the cost. Whoever knows this beforehand decides differently, and usually better.
Key takeaways
- The price on the invoice is only part of what software costs a company. The total cost of ownership model adds up all costs over the period of use.
- The largest items besides the license are almost always working time: setup, training, data migration, ongoing upkeep, and the time in which the old and new systems run in parallel.
- In the worked example of a CRM for eight users over three years, the license accounts for about 31 percent of total cost. Per user per month, the tool costs not 25 but just over 90 euros.
- Total cost is not an argument against software but the basis for an honest comparison with the benefit and with the alternative of changing nothing.
Why the license price is misleading about cost
Software for small companies is almost always rented today. An account is set up in minutes, the first fourteen days are free, and after that billing is per user per month. That lowers the hurdle, and it shifts the view. Whoever only sees the monthly price compares 19 with 29 euros and overlooks the items that appear on no price sheet: the week in which someone sets up the system, the hours of training, the months in which the old spreadsheet keeps running.
This leads to the guiding question of this guide: What does software really cost a small company over its entire period of use, and how can it calculate this before deciding, with reasonable effort? To answer it, we first clarify the terms and introduce total cost of ownership as a thinking model from purchasing. Then we review the research, weigh two perspectives, and work through a complete example that results in a cost sheet for your own decision.
A license buys the right to use a tool. It is paid for with the time of the people who are supposed to use it.
Terms: license, implementation, upkeep, switching
License costs are the price for the right of use, for rented software usually per user per month, often tiered into plans with different features. Implementation costs cover everything needed until the tool runs in daily work: setup, customization, data migration, training, and external support. Upkeep costs arise continuously: managing users and permissions, keeping templates and fields up to date, answering questions, keeping track of updates.
Two further items are almost always forgotten. Parallel costs arise as long as the old and new systems are maintained side by side. Switching costs arise at the end: exporting data, migrating it to a new system, changing processes. They arise only years later but are determined at selection, for example by whether all data can be exported in an open format. The article on project management tools compared names the same four items besides the license: implementation time, upkeep, switching costs, and duplicate systems.
The thinking model: total cost instead of purchase price
The thinking model comes from purchasing and is called total cost of ownership. It states that the price on the invoice is only part of what an item costs the company. Lisa Ellram (1995) described it as a procedure for buyers: all costs arising before the purchase, at the purchase, and during use until the end are recorded, using the same grid for each supplier. The article on reshoring or offshoring uses the same model for suppliers abroad.
An everyday image fits software. A cheap printer costs fifty euros; the cartridges cost many times that over the years. Nobody would compare two printers by purchase price alone if they knew how much they print. With software, the cartridges are your own people’s working hours, and they differ in amount with every provider.
| Phase | Item | How to estimate it |
|---|---|---|
| before purchase | selection and trial | hours of those involved in requirements and trial phase |
| implementation | setup, data migration, training, external help | hours per task times cost rate, plus quotes from external partners |
| use | licenses, plan changes, add-on modules | price per user times users times months, with expected growth |
| use | upkeep and support within the team | hours per week of a responsible person |
| transition | parallel operation | extra time per person per week until the old system ends |
| end | export, migration, archiving | hours for the exit, depending on open data formats |
Source: after the total cost of ownership approach in Ellram (1995)
What research and case studies show about software costs
That implementations turn out more expensive and longer than planned is well documented for larger systems. Iskanius (2009) studied the risks of projects introducing enterprise resource planning software in three manufacturing companies. She summarizes earlier surveys as showing that fewer than 30 percent of such implementations were completed on time, within budget, and with all required features. Her recommendation for small and medium-sized companies: break large implementations into manageable subprojects whose risks can be assessed individually.
Which costs are overlooked in daily work is described by Arum and Abdurrahman (2026) in a case study of a real estate office in Indonesia, whose IT costs they broke down for 2022 to 2024 using the total cost approach. Software was the largest item there. In addition, they found hidden costs from downtime, dependency on individual vendors, and troubleshooting time, which did not appear in the previous view based only on purchase price. For small companies, an older finding described in the article on choosing software also matters: in 114 small companies, the success of implemented systems depended strongly on the quality of external partners (Thong, Yap, and Raman 1996).
Source: Iskanius (2009); Arum and Abdurrahman (2026); Thong, Yap, and Raman (1996)
Rent cheaply or calculate carefully? Two perspectives
The first view says: for rented software, a total cost calculation is hardly worthwhile. The risk is small because you can cancel monthly, and a wrong choice costs a few monthly fees at most. Better to try than to calculate. This view is supported by the fact that the hurdle for a trial really is low.
The second view says: precisely because the license is cheap, the rest is overlooked. The license can be canceled, but not the hours that went into setup, training, and data migration. And the longer a tool runs, the more expensive switching becomes, because data, templates, and habits depend on it. Whoever only compares monthly fees ties themselves down without noticing.
I consider the second view the more robust one, with the justified core of the first. For a small tool used by one person, no calculation is worthwhile. For anything several people use daily or where customer and order data are stored, a simple one-page calculation pays off. It takes an hour and protects against the most expensive kind of bad purchase: the tool that was cheap, tied up many hours, and is gone again after a year.
Comparison by monthly price
- price per user in the cheapest plan
- free trial period
- number of features
- monthly cancellation possible
Comparison by total cost
- price in the plan that contains the needed features
- hours for setup, training, and data
- upkeep effort per week
- cost of exit, open data formats
Worked example: what a CRM for eight users costs over three years
A company introduces a CRM for eight users. All figures are freely chosen and rounded. The license costs 25 euros per user per month; from the second year, a higher plan at 30 euros becomes due for a needed add-on module. Internal hours are valued at 51 euros, the capacity cost rate from the article on what a process costs. The calculation covers three years of 46 working weeks each.
| Item | Assumption | Cost |
|---|---|---|
| licenses | 8 users × €25 × 36 months | €7,200 |
| plan change from year 2 | 8 users × €5 × 24 months | €960 |
| setup and training | 40 hrs. setup, 6 hrs. training per user, 88 hrs. in total | €4,488 |
| data migration | 20 hrs. | €1,020 |
| external support | quote from a partner | €2,000 |
| upkeep | 1 hr. per week over 138 weeks | €7,038 |
| parallel operation | 8 users × 0.5 hrs. per week × 13 weeks | €2,652 |
| exit at the end | export and migration, 20 hrs. | €1,020 |
| total | · | €26,378 |
Worked example
Worked example with freely chosen values. Licenses account for about 31 percent of total cost. Per user per month, the tool costs about 92 euros instead of 25 euros.
Two results are typical. First, upkeep is almost as expensive as the licenses, even though it amounts to only one hour a week. Second, implementation determines a large part of the cost: if the old system runs in parallel not for 13 but for 26 weeks, parallel costs rise to over 5,000 euros. What an implementation with a fixed cutover date looks like is described in the guide to introducing a CRM.
Total cost is half of the calculation. The other half is the benefit. If the CRM saves each of the eight people half an hour a week, for example through less searching and asking around, that is 552 hours or just over 28,000 euros over three years. The tool then just pays for itself. If it saves only a quarter of an hour, it does not pay for itself unless it brings additional revenue through fewer forgotten follow-ups. This calculation forces you to name the expected benefit in advance.
The cost sheet: total cost in five steps
Set period of use and users
Calculate with three years and the number of users you expect in a year, not only today’s.
Choose the right plan
Check which plan contains the features you will need in a year, and calculate with that price.
Estimate implementation in hours
Estimate setup, data migration, and training per person in hours and value them at your cost rate. Get a quote for external help.
Add ongoing and hidden items
Set an upkeep time per week, a duration for parallel operation with a fixed cutover date, and the hours for a later exit.
Compare with the benefit
Write down how many hours or what revenue the tool should bring per week, and check after three months whether that happens.
The cost sheet belongs to selection, not to the end. In the process from the guide to choosing software, it complements the weighted evaluation: two candidates with a similar score often differ markedly in setup effort and exit costs. For existing tools, the same calculation is worthwhile once a year. The four-question inventory from the article on reducing tool sprawl shows which subscriptions are barely used and which overlap.
- Is all data available in an open format that can be exported without the provider’s help?
- Is the calculation based on the plan containing the features needed in a year?
- Is the cutover date set on which the old tool ends?
- Is a person named with fixed weekly time for upkeep?
- Is the expected benefit written down in hours or euros?
Conclusion: what does software really cost?
The question was software costs a small company over its entire period of use and how it can calculate this in advance. According to the total cost model, implementation, ongoing upkeep, parallel operation, and the later exit belong alongside the license. Almost all of these items are working time. In the worked example, licenses account for only about a third of the cost, and the tool costs not 25 but just over 90 euros per user per month.
Research shows that implementations in particular turn out more expensive than planned and that partners and process determine costs more than the product. The calculation is therefore not an argument against software but for an honest decision: with costs and benefits on the same page. The sober observation is: software rarely becomes expensive through its price, but through the hours nobody counted beforehand.
How to get started
- Choose the tool that most people in your company use daily.
- Estimate the upkeep time per week for this tool and extrapolate it to a year at your cost rate.
- Check whether you are on the right plan and how many paid accounts are actually used.
- Do a test export of all data and note how long it takes.
- Set up the five-step cost sheet for your next software decision.
Frequently asked questions
What does total cost of ownership mean for software?
Total cost of ownership covers all costs of a piece of software over its period of use: licenses, implementation, training, data migration, upkeep, parallel operation, and exit. The monthly price is only part of it.
How do you calculate the cost of a software implementation?
Estimate the hours for setup, data migration, and training per person and multiply them by your internal cost rate. Add quotes from external partners and the time during which the old and new systems run in parallel.
What hidden costs does software have?
The most common are ongoing upkeep, plan changes as feature needs grow, parallel operation with the old system, downtime, and the cost of a later switch. Almost all of them consist of working time.
Is a total cost calculation worthwhile for subscription software?
For a small tool used by one person, usually not. For software that several people use daily or that holds customer and order data, a one-page calculation pays off. It takes about an hour.
How much does a CRM really cost?
That depends on the number of users, plan, and implementation. In a worked example with eight users and a license of 25 euros per user per month, total cost over three years was about 26,000 euros, or just over 90 euros per user per month.
Sources
- Ellram, L. M. (1995): Total cost of ownership: an analysis approach for purchasing. International Journal of Physical Distribution & Logistics Management, 25(8), 4-23.
- Iskanius, P. (2009): Risk Management in ERP Project in the Context of SMEs. Engineering Letters, 17(4), 266-273.
- Arum, T. A. S.; Abdurrahman, L. (2026): Optimizing Information Technology Investment Using the Total Cost of Ownership (TCO) Approach: A Case Study of a Property Company. Eduvest: Journal of Universal Studies, 6(6), 6630-6641.
- Thong, J. Y. L.; Yap, C.-S.; Raman, K. S. (1996): Top Management Support, External Expertise and Information Systems Implementation in Small Businesses. Information Systems Research, 7(2), 248-267.
