How to Build a Salesforce Implementation Business Case and ROI Model

Core decision: prove that a defined operating problem is worth solving, Salesforce is the best available option, and risk-adjusted benefits exceed full lifecycle cost.

Core modeling rule: use your own baseline, adoption ramp, gross-margin economics, cashability, attribution, recurring cost, and delivery risk.

Cost input: build the full Salesforce implementation cost before calculating return. A service quote is not the total cost of ownership.

Approval rule: approve only when the base case clears the finance hurdle, the downside is acceptable, benefit owners are named, and measurement is funded after go-live.

The Business Case at a Glance

Decision question	Evidence required
Why change?	A measured business problem, its cost, and the consequence of doing nothing.
Why Salesforce?	A capability-to-outcome chain and comparison with credible alternatives.
What will it cost?	One-time investment, recurring run rate, internal effort, contingency, and end-of-life cost.
What will it return?	Risk-adjusted annual benefits, cash flow, ROI, NPV, payback, and break-even points.
What could stop value?	Delivery, data, adoption, integration, security, and operating-model risks with owners.
Who is accountable?	A named executive sponsor, product owner, finance reviewer, and benefit owner for each KPI.

A credible business case is not a feature list. Australian Government investment guidance uses the same decision logic: define the problem, compare options, estimate costs and benefits, expose assumptions, and communicate risks.

How to Build the Business Case

1 Define the Decision

• Decision statement: Approve [release] for [users/processes] by [date] at a lifecycle cost not exceeding [amount], subject to [gates].
• Business problem: Name the current failure in operating terms, not as a missing Salesforce feature.
• Release boundary: List products, processes, teams, regions, data, integrations, reports, and explicit exclusions.
• Evaluation horizon: Set the currency, price basis, model start date, contract term, and the number of years covered.
• Decision criteria: Agree on ROI, NPV, payback, affordability, compliance, and strategic-fit thresholds before final estimates are presented.

2 Establish the Baseline

• Use system data, finance records, time studies, and process sampling. Interviews explain the problem but should not be the only baseline.
• Use a full sales or service cycle when possible. Use 12 months for seasonal metrics. Record source, owner, period, exclusions, and data-quality limits.

Baseline metric	Definition	Primary evidence
Lead conversion	Converted leads / qualified leads	CRM export and finance-accepted stage definitions
Win rate	Closed-won opportunities / decided opportunities	Opportunity history by segment
Sales cycle	Median days from qualified opportunity to close	Stage-change timestamps
Seller administration	Hours per user per week on entry, routing, reporting, and reconciliation	Time study plus activity logs
Forecast error	Absolute variance between forecast and actual	Forecast snapshots and booked revenue
Service cost	Fully loaded service cost / resolved cases	Workforce, telephony, and finance data
Resolution time	Median time from case open to resolution	Case timestamps by priority
Retention	Customers or recurring value retained / renewable base	Billing, contract, and account data
Data rework	Hours or cost spent correcting duplicates, missing fields, and sync failures	Ticket, audit, and operations logs
Technology run rate	Annual cost of systems that can actually be retired	Contracts, invoices, and exit terms

3 Build the Outcome Chain

• Required chain: Salesforce capability -> changed user behavior or process -> operational KPI -> financial effect.
• Reject any proposed benefit that cannot be connected through the full chain or assigned to a measurable baseline.

Capability	Operating change	KPI	Financial effect
Automated lead routing	Qualified leads reach the correct owner faster	Response time; conversion	Incremental gross profit
Guided selling and approvals	Fewer handoffs and pricing delays	Cycle time; approval rework	Capacity value; earlier cash flow
Unified customer record	Less searching and duplicate entry	Admin hours; duplicate rate	Cashable labor or avoided rework
Service automation	Cases are classified, routed, and resolved with fewer touches	Handle time; cost per case	Avoided hiring, overtime, or vendor spend
Self-service	Eligible contacts complete tasks without an agent	Deflection; completion; repeat contacts	Cost-to-serve reduction
System consolidation	Legacy tools and interfaces are retired	Applications retired; support hours	License and maintenance savings

4 Compare Credible Options

• A business case should prove that Salesforce is the preferred option, not merely that the proposed project has benefits.

Option	What to estimate	When it may win
Do nothing	Failure cost, risk exposure, manual effort, lost revenue, and current run rate	The problem is small, temporary, or cheaper to tolerate
Optimize current stack	Process redesign, configuration, integration, training, and remaining constraints	Most value can be recovered without platform replacement
Phased Salesforce release	Release-one cost and benefits, later options, dependencies, and rework risk	Evidence is incomplete or early value can fund later scope
Full transformation	Multi-year platform, migration, integration, change, and operating-model cost	Cross-functional value materially exceeds complexity and risk

5 Calculate Full Lifecycle Cost

• Use the parent Salesforce implementation cost guide to build like-for-like Year 1 and run-rate inputs. Then extend those inputs across the ROI horizon.
• Cost-quality test: the GAO Cost Estimating and Assessment Guide says reliable estimates require a technical baseline, assumptions, data, methodology, sensitivity and risk analysis, documentation, and updates with actual cost.

Cost category	Include
One-time delivery	Discovery, architecture, configuration, development, testing, deployment, and hypercare
Data	Profiling, cleansing, mapping, migration cycles, reconciliation, archiving, and retention
Integration	Connectors, middleware, APIs, identity, error handling, monitoring, and support handoff
Licenses and usage	Salesforce editions, add-ons, sandboxes, external users, data, AI, storage, and consumption
Internal effort	Product owner, SMEs, data owners, security, finance, UAT, training, communications, and backfill
Change and adoption	Role design, process change, learning, champions, office hours, and adoption analytics
Run rate	Administration, support, release management, integration operations, data quality, and enhancements
Transition and exit	Parallel running, contract termination, decommissioning, archive, and end-of-life work
Risk reserve	Named unresolved risks, not an unexplained percentage added to every estimate

6 Build the Benefits Register

• Keep cash-releasing, capacity, risk, and qualitative benefits separate. Only cash-releasing and defensibly monetized benefits belong in headline ROI.
• Give every benefit one owner, one baseline, one formula, one data source, one realization date, and one anti-double-counting rule.

Benefit	Calculation	Evidence and control
Labor capacity	Eligible users x hours saved x loaded hourly cost x adoption x cashability	Time study; do not count 100% unless overtime, hiring, contractor spend, or measurable output changes
Incremental sales	Eligible volume x conversion lift x average value x gross margin x attribution	Cohort or controlled comparison; use gross profit, not booked revenue
Retention	Renewable base x churn reduction x contribution margin x attribution	Comparable customer cohorts; exclude unrelated pricing or product effects
Service savings	Volume x unit-cost reduction, or avoided FTE/vendor cost	Case volume, cost per resolution, service level, and quality guardrails
Technology savings	Retired license + maintenance + support – exit and replacement cost	Executed retirement plan and contract dates
Working capital	Cash-flow timing benefit from faster billing, collections, or fewer disputes	Finance-approved method; do not count the full receivable balance as profit
Risk reduction	Annual event probability x financial impact x expected reduction	Risk owner, incident history, control evidence, and conservative probability
Qualitative value	Score or threshold, not invented currency	Compliance, employee experience, customer visibility, or strategic optionality tracked separately

7 Build the ROI Model

Measure	Formula	Use
Net cash flow	Risk-adjusted benefits – one-time cost – recurring cost	Year-by-year value
Three-year ROI	(Total benefits – total lifecycle cost) / total lifecycle cost x 100	Simple return over the decision horizon
NPV	Sum of net cash flow in year t / (1 + discount rate)^t	Time-adjusted value using the finance-approved hurdle or discount rate
Payback	First month when cumulative net cash flow becomes zero or positive	Liquidity and recovery speed
IRR	Discount rate that makes NPV equal zero	Comparison with the corporate hurdle rate; use with NPV, not alone
Break-even	Value of a key assumption that makes NPV or ROI equal zero	Decision resilience

• Discounting rule: use one consistent real or nominal basis. OMB benefit-cost guidance explains that future benefits and costs should be discounted and that incremental effects should be separated from past costs and double counting.
• Do not copy a public-sector discount rate: use the rate approved by your finance team. The worked example below uses 10% only to demonstrate the calculation.

8 Risk Adjust and Stress Test

• The UK Green Book recommends explicit adjustments for optimism bias by increasing expected costs and durations and decreasing expected benefits, followed by sensitivity and switching-value analysis.
• Create conservative, base, and upside cases. Change linked assumptions, not the final ROI percentage.

Variable to test	Downside question
Adoption	What if active use reaches only 60% of eligible users?
Benefit ramp	What if value begins one or two quarters later?
Cashability	What if saved time does not reduce spend or create measured output?
Revenue attribution	What if Salesforce receives only 25% to 50% of the observed uplift?
Cost overrun	What if data, integration, security, or testing costs rise 15% to 30%?
Run rate	What if support, consumption, storage, or enhancement demand is higher?
Scope and schedule	What if release one is narrower or go-live moves by one quarter?
Benefit durability	What if improvement decays after the first year without optimization?

9 Assign Owners and Measurement

• The PMI benefits realization guide connects strategy, project deliverables, benefits, and success measurement. The project cannot own benefits after the operating team takes over.

Role	Accountability
Executive sponsor	Owns the decision, strategic outcome, and cross-functional tradeoffs
Finance	Validates unit economics, cashability, attribution, discount rate, and reported value
Product owner	Owns release scope, adoption, backlog, and operating-model change
Benefit owner	Owns one KPI, target, data source, realization date, and corrective action
Process and data owners	Own definitions, baseline quality, controls, and sustained process compliance
Technical lead	Validates feasibility, architecture, non-functional scope, estimates, and technical risk
PMO or value office	Maintains assumptions, gates, benefits register, actuals, and variance reporting

Worked Three Year Salesforce ROI Model

• Status: hypothetical planning example in USD, not a benchmark, quote, or expected return.
• Scope: 100 sales users, 25 service users, one core release, selected automation, legacy-system retirement, and three-year operation.
• Method: benefits are reduced for adoption, attribution, and cashability before they enter the model.

Lifecycle Cost

Cost input	Timing	Amount
Configuration, development, and testing	Before go-live	$150,000
Data migration and integrations	Before go-live	$60,000
Internal SMEs, UAT, and backfill	Before go-live	$35,000
Training and change	Before go-live	$20,000
Contingency for named risks	Before go-live	$25,000
Licenses and apps	Each operating year	$150,000
Administration and support	Each operating year	$60,000
Release and optimization capacity	Each operating year	$30,000
Three-year lifecycle cost	Year 0 through Year 3	$1,010,000

Full Run Rate Benefits

Benefit input	Calculation	Annual value
Sales capacity	100 x 1.5 hours/week x 52 x $60 x 60% cashability	$280,800
Service capacity	25 x 2 hours/week x 52 x $40 x 50% cashability	$52,000
Incremental gross profit	2,000 opportunities x 1 percentage-point lift x $20,000 x 45% margin x 60% attribution	$108,000
Retired technology	Contracts and support removed after exit costs	$90,000
Avoided errors and rework	Finance-validated reduction in annual operating cost	$60,000
Full annual benefit	Before the adoption ramp	$590,800

Cash Flow and Return

Period	Benefit ramp	Benefits	Costs	Net cash flow	Cumulative
Year 0	0%	$0	$290,000	-$290,000	-$290,000
Year 1	50%	$295,400	$240,000	$55,400	-$234,600
Year 2	80%	$472,640	$240,000	$232,640	-$1,960
Year 3	90%	$531,720	$240,000	$291,720	$289,760

• Three-year ROI: 28.7%.
• NPV at an illustrative 10% discount rate: $171,802.
• Payback: about 24.1 months after go-live.
• IRR: about 34.4%.
• Robustness check: removing all modeled revenue uplift still leaves a 5.2% undiscounted three-year ROI. The case is not dependent on speculative sales growth.

Scenario Analysis

Scenario	Changed assumptions	Three-year ROI	NPV at 10%	Decision signal
Conservative	Benefits 25% below base; initial cost 15% above base	-7.5%	-$136,360	Do not approve unchanged; reduce scope, cost, or risk
Base	Modeled assumptions	28.7%	$171,802	Compare with hurdle rate and risk appetite
Upside	Benefits 15% above base; initial cost 5% below base	50.1%	$345,099	Useful for capacity planning, not the approval case

• Switching value: base-case benefits can fall about 22.3% before undiscounted three-year ROI reaches zero.
• Decision insight: the conservative case fails. Approval should therefore include adoption, retirement, and cost-control gates rather than rely on the base-case percentage alone.

Real Company Results and How to Use Them

• Use published customer outcomes to identify plausible benefit categories and measurement methods. Do not import another company’s ROI, adoption rate, scope, cost base, or attribution into your model.
• The examples below are vendor-published customer stories. They are useful evidence of possible mechanisms, but they are not independently audited benchmarks or forecasts for your organization.

Company	Published result	What it can inform	What not to assume
Year Up	45% higher matching efficiency and about 4,500 staff hours saved annually; enrollment administration time also fell 42%.	Time-study design, process baseline, and capacity-value formula	That every saved hour becomes payroll savings
T-Mobile	70% reduction in sales work effort for a custom app that let representatives place orders in minutes.	Task-level effort baseline and workflow simplification	That the same reduction applies to every sales process
FedEx	13% improvement in customer activation and more than 2,000% reported ROI from unifying sales, shipping, and web data for dormant-account action.	Activation cohorts, data-unification benefits, and revenue attribution	That an outlier ROI is a suitable planning benchmark
Sutton Tools	more than 100 reporting hours saved monthly, 70% fewer inbound calls, and 4x online orders within 12 months.	Separate productivity, service, and channel-adoption benefits	That all observed commerce growth was caused by Salesforce alone

What Research and Experts Say

Expert evidence	Business case implication
“Organizational ‘investments’ have a large influence on the value of IT investments.” – Erik Brynjolfsson and Lorin Hitt	Budget process redesign, skills, data ownership, management time, and adoption. Software cost alone cannot explain return.
“CRM must be conceived as a strategy, due to its human, technological, and processes implications.” – Luis Mendoza and co-authors	Test people, process, and technology assumptions together. A feature-only case is incomplete.
“Firms show improvements in operational performance, operational efficiency, accounts receivable collectability, and earnings predictability.” – Jacob Haislip and Vernon Richardson	The benefit register can extend beyond sales uplift, but each category still needs a company-specific baseline and attribution rule.
Research on the productivity J-curve shows that new technology often requires complementary investment before benefits are harvested.	Use an adoption ramp. Do not place full benefits in the first month after go-live.

 

Common ROI Errors

• Starting with a desired ROI and reverse-engineering the assumptions.
• Using vendor averages or customer stories instead of the organization’s baseline.
• Counting revenue instead of incremental gross profit.
• Valuing every saved hour as cash without a redeployment, overtime, contractor, or hiring plan.
• Counting faster sales cycles, higher conversion, and more revenue as three separate benefits when they overlap.
• Applying full benefits from go-live with no learning, adoption, data-quality, or stabilization ramp.
• Excluding licenses, internal labor, integration operations, support, enhancements, or decommissioning.
• Treating sunk cost as a reason to continue; the decision should use future incremental costs and benefits.
• Reporting one precise ROI with no scenario, range, switching value, or assumption confidence.
• Ending governance at deployment instead of tracking benefits against finance data after go-live.

Business Case Approval Pack

Artifact	Minimum content	Approval question
Executive summary	Decision, problem, preferred option, investment, return, risks, and gates	What exactly are we approving?
Scope and architecture	Release boundary, integrations, data, security, environments, and exclusions	Is the estimate tied to a real solution?
Baseline pack	Metric definitions, periods, sources, owners, and data-quality limits	Can finance reproduce the starting point?
Options analysis	Do nothing, optimize, phase, and transform where credible	Why is this option preferred?
Cost model	Year 1, run rate, internal effort, contingency, and exit	Is the investment affordable and complete?
Benefits register	Formula, source, ramp, attribution, cashability, owner, and target date	Can each benefit be defended and owned?
Financial model	Cash flow, ROI, NPV, payback, IRR, and break-even	Does value clear the agreed thresholds?
Risk and scenario pack	Downside, base, upside, switching values, mitigations, and reserves	What happens when assumptions miss?
Realization plan	30-, 60-, 90-day and quarterly measurement with corrective actions	How will value be proven after go-live?

Benefits Tracking After Approval

Timing	Measure	Action
Before build	Baseline, segment definitions, data quality, and control groups where feasible	Freeze definitions and obtain finance sign-off
At go-live	Deployment cost, scope variance, readiness, and starting adoption	Reset forecasts for approved changes
30 days	Access, activity, defects, data completeness, and training completion	Remove adoption blockers
60 days	Workflow compliance, cycle time, automation success, and support demand	Correct process and configuration issues
90 days	Early operational KPI movement and cost run rate	Validate or revise the ramp
Quarterly	Realized cash, capacity, revenue, risk, and qualitative benefits	Finance validates value and owners commit corrective action
Annually	TCO, license utilization, retirement, benefit durability, and next-release value	Refresh the roadmap and stop low-value scope

Frequently Asked Questions

What is a good ROI for a Salesforce implementation

• There is no universal percentage. A good result clears the organization’s finance hurdle, beats credible alternatives, remains acceptable in the downside case, and does not depend on weak revenue attribution.

How many years should the ROI model cover

• Use the period needed to capture material implementation, operating, renewal, benefit, and exit effects. Three years is a clear initial view for many software decisions; extend it when contracts, migration, or benefits materially continue beyond that point.

Should ROI include Salesforce licenses

• Yes. Include all incremental licenses, add-ons, usage, apps, middleware, support, administration, and optimization required to sustain the modeled benefits.

Can time savings be counted as ROI

• Only the cashable or measured productive share. Link saved time to avoided hiring, lower overtime, reduced contractor spend, increased throughput, or another finance-accepted outcome. Track the remainder as capacity, not cash.

How should revenue uplift be calculated

• Use eligible volume x incremental change x average value x contribution or gross margin x Salesforce attribution. Exclude growth caused by price, market demand, new products, staffing, or campaigns unless the model controls for them.

What if an important benefit cannot be monetized

• Quantify it where possible, keep it outside headline ROI, and show the break-even value it would need to justify the decision. Do not invent a dollar value for visibility, morale, compliance, or customer experience.

Who should own Salesforce ROI

• The executive sponsor owns the investment decision, finance validates the economics, and operational benefit owners own the KPI movement. IT or the implementation team should not own every business benefit.

When should the business case be updated

• Update it after discovery, before contract commitment, after material scope or schedule changes, at go-live, and during quarterly benefits reviews. Replace estimates with actual cost and performance data as soon as it is available.

Final Takeaway

• Start with a measured problem and compare credible options.
• Model full lifecycle cost, not only implementation services.
• Convert benefits with adoption, cashability, gross margin, attribution, timing, and risk.
• Present ROI with NPV, payback, scenarios, and switching values.
• Assign operational owners and keep measuring after go-live.
• Before presenting the return, verify every cost input against the Salesforce implementation cost model.