SNF Software Total Cost: A Five-Year Model With Pricing and Staffing Tradeoffs

Build a model the CFO can reproduce

Set a start date, planning horizon, facility list, billing units and price basis. Compare proposals using identical assumptions. An annual subscription, a monthly connection fee and one-time migration services belong in different rows. State whether tax, travel and third-party charges are included. The GAO Cost Estimating and Assessment Guide supports explicit scope, assumptions, sensitivity analysis and updates with actual costs; the model below is an original SNF purchasing example.

Do not begin with a promised percentage improvement. First determine the spending and capacity required to operate the proposed arrangement, including the work between vendors and the facility.

Work the five-year cash model

All amounts below are hypothetical, with no inflation, discounting or tax adjustment. Subscription and interface support are paid for each full operating year.

Exit cost here assumes a transition at the end of year five. A renew-and-continue scenario needs a different terminal assumption. Do not show the $10,000 as a certain future invoice merely because it occupies a row.

Suppose subscription increases 10% once at the start of year three. Add $6,000 in each of years three through five: $18,000. A new interface in year two adds $5,000 setup and $2,000 support for four years: $13,000. The changed cash scenario is $442,000. A $5,000 implementation discount would not offset these $31,000 of changes.

The decision question is which commercial terms control each assumption. ONC EHR Contracts Untangled, September 2016 discusses integration and transition arrangements as procurement issues. Confirm included work and change triggers against the actual agreement rather than assuming a standard market practice.

Compare pricing units before comparing unit prices

Consider two fictional subscription formulas, both with the same $400 monthly base fee:

• Bed formula: $12 × billable beds + $400.
• User formula: $9 × billable users + $400.

At 100 beds and 140 users, the monthly amounts are $1,600 and $1,660. Without minimums, they cross where users = 12 ÷ 9 × beds: approximately 133.3 users for 100 beds. This is an arithmetic boundary, not a recommended staffing ratio.

If the bed proposal imposes a 120-bed minimum, its charge becomes $1,840 even at 100 beds. The user formula crosses that amount at 160 users. The minimum changes the preferred price mechanically. Before using either result, establish whether beds means contracted, licensed or occupied beds and whether users means named, active or concurrent accounts. Agency accounts and regional staff working at multiple facilities can materially change the billing population.

A lower subscription does not win the full cost comparison if it excludes required modules, support or interfaces. Put the normalized subscription back into the five-year model.

Budget staff capacity without counting it twice

Estimate migration review from work, not a percentage of subscription cost. An illustrative sample has 120 source exceptions at 10 minutes each, six two-person workshops lasting two hours, 80 record reviews at 15 minutes and 40 retests at 15 minutes. That is 20 + 24 + 20 + 10 = 74 labor-hours. Meeting duration alone would understate the workshops by half.

If source exceptions double, add 20 hours; if a qualified reviewer is unavailable, the elapsed schedule can slip without increasing review hours. Keep waiting and labor separate.

For a broader 600-hour implementation at a planning value of $40/hour, report $24,000 of internal capacity. If 100 of those hours require $50/hour replacement coverage, show $5,000 incremental cash. Do not add $24,000 and $5,000 into an unexplained “labor cost”: one values existing staff work and the other pays for coverage. Identify the accounting view and any overlap.

An independent facility may have one employee performing mapping, validation and training, making simultaneous work impossible. A group should budget common mapping once and local validation per site; multiplying one site's complete estimate by the facility count can misstate both.

Test the benefit against recurring review work

A fictional task occurs 300 times weekly. Only 240 routine cases improve by two minutes; 60 exceptions are unchanged. Gross released capacity is eight hours, not ten. If the new workflow adds two weekly hours of reconciliation, net modeled capacity is six hours.

At a hypothetical $40/hour over 50 weeks, that capacity has a planning value of $12,000 annually. It does not by itself offset an additional $20,000 cash subscription. To cover that cash amount at $40/hour over 50 weeks would require ten hours per week of actual avoidable expenditure. Reject a business case that silently treats fragmented minutes as payroll savings.

Maintain conservative and higher-work scenarios, with named assumptions rather than an unexplained contingency. After selection, replace estimated unit counts, coverage payments and reconciliation time with actuals. The model should expose why the decision changes, not merely produce a precise-looking total.