“The build is a one-off. After that, it's just hosting.”
Hana puts the proposal beside last year's support notes at Sandline. Rule changes, training and a connection nobody remembers ordering. “Just hosting” has been busy.
Her real question is simple: what will this system cost to own?
Hana is comparing keeping, rebuilding and replacing it over five years. Let's count the work, then see what could change her decision.
What does total cost of ownership mean?
Total cost of ownership, or TCO, is the cost across the time you use something. For software, count buying, running, changing and leaving it.
The UK government's TCO checklist includes buying, maintenance, training, changes and end-of-life costs. It covers financial cost only. Check fit, reliability and risk on their own.
Beezr's Ownership Iceberg makes that bill visible. Build or licence costs and yearly fees sit above water. Support, changes, training, integrations, governance and exit sit below.
This is Beezr's model, not an accounting standard. “Below” means you must ask about it. It needn't be the larger part: your inputs set the split.
What goes in the total cost of ownership calculator?
Give each option the same job. A quote tool and a full order process aren't the same scope. Neat columns won't fix that.
Name up to three options. The starting names are “Keep the current system”, “Build custom” and “Configure a platform”. All start with identical values, labelled illustrative: replace with your numbers.
Load Hana's example to use her fictional estimates below.
| Group | Input | How it is counted |
|---|---|---|
| Above the waterline | Upfront: licences, build or implementation | Once at the start |
| Above the waterline | Yearly subscription or hosting | Every year |
| Below the waterline | Support and maintenance, including people and vendor support | Every year |
| Below the waterline | Changes per year × average cost per change | Every year |
| Below the waterline | Initial training + yearly refresh | Once, plus every year |
| Below the waterline | Number of integrations × build cost | Once at the start |
| Below the waterline | Number of integrations × yearly upkeep | Every year |
| Below the waterline | Governance and compliance effort, such as access reviews and audits | Every year |
| Below the waterline | Exit migration cost | Once at the end, if enabled |
Count staff time in the right input, too. If Karim maintains a connection, estimate his work the same way throughout. Existing staff time isn't free in just one column.
Count each task once. Split out integration upkeep if support already includes it.
Note what you don't know beside the estimate. Zero means no cost expected, not “we haven't asked”.
What is the total cost of ownership formula?
Choose 1–7 years, default 5. Select AED, EGP, SAR or USD for display.
The selector doesn't convert money. Use one currency for all inputs.
TCO = upfront + training_initial + integrations × build_cost
+ Σ(years) [subscription + support + changes × change_cost
+ training_refresh + integrations × upkeep + governance]
+ exit
The sum repeats the annual bracket for each year. With unchanged annual inputs, multiply that bracket by the selected number of years. exit is zero when exit is off; otherwise, add it once at the end.
Above water = upfront + subscription × years. Below = everything else. Training and integration builds stay below, though paid at the start. The split shows visibility, not when bills fall due.
Name up to three options. Add upfront, initial training and integration build costs. Add each year's subscription, support, changes, training refresh, integration upkeep and governance. Add exit once at the end if enabled.
Choose 1–7 years, default 5. Currency changes display only.
Above water is upfront plus subscription multiplied by years. Everything else is below. These are estimates, not quotes or a recommended winner.
This simple estimate doesn't discount future costs. It leaves out inflation, financing, taxes, exchange movements and money's changing value over time. It doesn't price the chance or impact of outages or security incidents.
Discuss those risks on their own. Still count the daily work you can estimate.
What does Hana's five-year comparison show?
Sandline · example
Hana gives each option the same job
Each option must handle the same process, approvals and connections. These fictional AED estimates cover five years, with exit on. They're illustrative assumptions, not Beezr prices, supplier quotes or market benchmarks.
Hana expects six rule changes a year and two integrations each. She imagined different costs for each option. No option is cheaper by nature.
| Input | Keep the current system | Build custom | Configure a platform |
|---|---|---|---|
| Upfront | 10,000 | 60,000 | 30,000 |
| Subscription or hosting / year | 12,000 | 6,000 | 18,000 |
| Support / year | 18,000 | 12,000 | 10,000 |
| Changes / year | 6 | 6 | 6 |
| Average cost / change | 1,000 | 1,500 | 800 |
| Initial training | 2,000 | 4,000 | 4,000 |
| Training refresh / year | 1,000 | 1,000 | 2,000 |
| Integrations | 2 | 2 | 2 |
| Build cost / integration | 3,000 | 5,000 | 4,000 |
| Upkeep / integration / year | 1,000 | 1,500 | 1,000 |
| Governance / year | 4,000 | 4,000 | 5,000 |
| Exit at end | 10,000 | 8,000 | 12,000 |
For “Keep the current system”, starting costs are 10,000 + 2,000 + (2 × 3,000) = 18,000.
Annual costs are 12,000 + 18,000 + (6 × 1,000) + 1,000 + (2 × 1,000) + 4,000 = 43,000.
Five years: 18,000 + (5 × 43,000) + 10,000 = 243,000. Above water: 10,000 + (5 × 12,000) = 70,000. Below: 243,000 − 70,000 = 173,000.
You can check each total against the inputs.
| Calculated result, AED | Keep the current system | Build custom | Configure a platform |
|---|---|---|---|
| Starting costs | 18,000 | 74,000 | 42,000 |
| Annual costs | 43,000 | 35,000 | 41,800 |
| Five-year total, including exit | 243,000 | 257,000 | 263,000 |
| Above the waterline | 70,000 | 90,000 | 120,000 |
| Below the waterline | 173,000 | 167,000 | 143,000 |
| Largest below-waterline item over five years | Support: 90,000 | Support: 60,000 | Support: 50,000 |
At six changes a year, keeping it costs least: 243,000 versus 257,000 and 263,000. That doesn't mean renew it. Hana still needs to check the agreed requirements.
Which guess could change the decision?
Double the current system's six changes a year to twelve. That adds 6 × 1,000 × 5 = 30,000. Its total reaches 273,000; the other two stay as they are.
The order flips. In this example, how often the business changes decides which costs least.
That doesn't prove there will be twelve changes. It's a sensitivity check: try another input and watch the effect. If it flips the order, research it.
Ask for a change log, recent support work and a typical change request. Those help refine the estimate.
Use the numbers to ask better buying questions
- Hana, Finance Agree the scope — same outcome and time span for each option.
- Karim, Systems Count the work — split initial, yearly and exit effort without overlap.
- Process owner Test the guesses — change the rate of rule changes; check the effect.
- Layla, Operations Check the fit — review needs and risks beside costs.
Government purchasing guidance asks teams to choose build, buy or a mix. Be realistic about skills and contract limits. What will your own people still need to do?
| Option | What to test | The catch to include |
|---|---|---|
| Keep the current system | Can it still meet the required process? | Familiarity doesn't settle the cost of future changes |
| Build custom | Can a maintainer demonstrate the difficult requirement? | Dedicated ownership continues after delivery |
| Configure a platform | Can the required rule be changed and reviewed using its components? | Fit, configuration skill and exit work still need evidence |
Agree what fits before comparing costs
- If an essential requirement isn't met → resolve the gap first — a low bill won't supply what's missing.
- If options meet the scope → compare lifetime costs — count people's work and exit on the same basis.
- If one guess flips the order → research it — check the driver before polishing the chart.
Myth / Reality
“The subscription is the total.” It's one input. Enter support, changes and staff work unless counted elsewhere. An included service belongs once.
Our take
Our take. Give owning software as much care as creating it. We'd rather see an honest open question than a tidy total built on convenient zeros.
How Beezr does it
Try a real change. In Beezr, Sandline's quote calculation can be a Smart Script. Its approval can be a workflow step. Sandbox changesets show item-level differences, conflict flags and rollback.
See the developer and security pages. Time a typical change, then estimate the review work. Use that to find your own change cost.
Key takeaways
- If you're technical: estimate real changes, integration upkeep and an exit you can rehearse. Count each task once.
- If you're not: ask which staff tasks are missing. Who supplied each guess?
- Check fit and risk beside costs.
- Totals and waterline splits describe your inputs. In Hana's example, the rate of change decides which costs least.
Try the process before estimating its upkeep
Pick one Beezr process and make the work concrete.
- Build records and forms — add modules, fields and relations.
- Set the changing rule — a Smart Script calculation and a workflow approval.
- Review the change — inspect the sandbox changeset's item-level differences.
- Check the exit — export any list to Excel; get a full SQL dump on request. List exports follow active filters, so check the scope.
Questions readers ask
What does TCO mean in procurement?
Buying costs plus ongoing ownership costs over an agreed time. Use the same needs and cost groups for each option. Check fit and risk separately.
Where did the TCO idea come from?
The UK government's checklist says Gartner first developed TCO models in 1987. We borrow the principle, not the rulebook.
Do all models use this total cost of ownership formula?
No. Ours adds initial costs, yearly costs and optional exit. Other models may add categories or discount future costs.
Can I use the total cost of ownership calculator for one year?
Yes: choose 1–7 years, default 5. Starting costs still count. At one year, add one year's costs and exit if enabled.
Does the calculator choose the best option?
No. It shows totals, the waterline split, cumulative costs and the largest item below water. Hana must still check needs, uncertainty and who owns the work.
Does a data export make exit cost zero?
Don't assume so. Check the files and what the receiving system needs. Use the worksheet in the data control guide.
Sandline Supplies is a fictional example company.
Sources
- Total Cost of Ownership: Things to Consider · UK Government, HMG ICT Strategy · 2011-10-27
- Define your purchasing strategy · Government Digital Service and Central Digital and Data Office · 2026-09-03