Imagine an illustrative 40-consultant firm that runs its pipeline in Salesforce and everything after the signature in spreadsheets. Sales closes a $120,000 fixed-fee engagement on Monday. By Wednesday a delivery lead builds a staffing plan from memory, and by month end finance is asking why the project billed 90 percent of its fee while consuming 115 percent of its hours. Nobody can see utilization across the firm without a manual export, and the partners argue about whose numbers are right. The CRM did its job, and the work that followed had no home.
This article explains what Salesforce does well for a services firm, where its native tools stop and how firms close the gap for projects, resource planning, utilization and billing. It draws on 3 sources written by firms with a commercial stake: a PSA vendor’s page, a software selection consultancy’s guide and an accounting software company’s utilization article that cites an SPI Research benchmark. Their figures are mostly unsourced, so each number is attributed. Confirm current features and prices with Salesforce and with any vendor you evaluate.
TL;DR
- The CRM is strong before the signature and thin after it. Kantata, a vendor of a competing project-based tool, says Salesforce handles pipeline and relationships well but wasn’t built to show skills, billable utilization or capacity on its own. Firms with several concurrent engagements usually add a professional services automation layer.
- Utilization needs a definition before it needs a dashboard. Billable utilization is billable hours divided by available hours, and available hours should exclude leave, holidays and training. A Beancount.io article cites SPI Research for a 66.4 percent average in 2025, which it calls the lowest in 19 years.
- Billing model drives system fit. Time and materials, fixed fee, retainer and milestone work each need different data and controls. In an illustrative example, a $120,000 fixed fee that takes 690 hours instead of 600 loses about 6.4 points of margin. Choose a tool that supports your main model natively. Ask for a live demo.
Where the CRM fits a services firm
It’s strong at winning the work. Salesforce tracks accounts, contacts, opportunities and forecasts, and a services firm lives on relationships and repeat business. Kantata’s page calls relationship and pipeline management a strength, and says Salesforce offers solid foundations such as tasks, activities and automation through Flow. A firm that sells complex engagements to the same clients year after year gets real value from knowing who the buyers are and what they’ve bought before.
It’s thin at running the work. The same page says Salesforce wasn’t built to provide visibility on skills, billable utilization or capacity on its own. It also names pipeline-to-capacity linking, revenue recognition tied to delivery milestones and real-time project margin as gaps. Those claims come from a vendor that sells a competing tool, so treat them as a hypothesis and test them against your own requirements, yet they match what many services firms report about CRM-only setups.
The handoff is where the pain shows up. When a deal closes, someone must turn it into a project: scope, budget, team, schedule and billing plan. If that step happens in email and spreadsheets, errors multiply. The sold scope drifts from the delivered scope, and finance bills from a single version while delivery works from another. A firm should decide early where the project record lives, and how the opportunity data flows into it.
Choose the system by the question you can’t answer today. If the problem is staffing, you need resource visibility. If it’s margin, you need project financials. If it’s cash, you need billing control. Write down the 3 or 4 questions leaders ask each week that nobody can answer quickly, and rank them. The ranking tells you whether a light fix, such as reports and Flow, will do, or whether you need a full professional services automation product.
What the CRM does natively and what it doesn’t
The table follows Kantata’s description of native capabilities and gaps, supplemented by the Viewpoint Analysis guide. Kantata sells a competing product and Viewpoint Analysis sells software selection services, so neither is neutral. Use the table to build a requirements checklist, and test each row in a demo of the tools you’re considering, using your own projects and rate cards. Check every row. Treat each row as a question for a demo.
| Need | Native Salesforce (per Kantata) | Typical gap |
| Pipeline and relationships | Strong | None for most firms |
| Tasks, activities and simple automation | Available through Flow | Limited project structure |
| Basic project tracking | Through AppExchange apps | Depth varies by app |
| Skills, utilization and capacity | Not provided on its own | Staffing is reactive |
| Pipeline-to-capacity view | Not provided | Hiring and bench decisions lag |
| Revenue recognition by milestone | Not provided | Manual schedules |
| Concurrent engagements and complex billing | Hard to run natively | Spreadsheet workarounds |
| Real-time project margin | Not provided | Margin erodes unseen |
Build options for projects and resources
- Native build with tasks, activities and Flow. Kantata’s page says this works for lightweight internal planning, and that the ceiling becomes clear quickly for firms with multiple concurrent engagements. A small firm with a few projects and a single billing model may be fine. Estimate the effort of maintaining your own objects, reports and automation, since the person who builds it becomes a dependency. Ask who will maintain it. Plan for upkeep.
- AppExchange project tools. The page mentions basic project tracking through AppExchange apps without evaluating specific products. Test any app against your real projects, including budgets, time entry and billing links. A tool that tracks tasks well may not track hours against fees, and the gap appears at month end when finance needs numbers. Test hours against fees in the trial, and check that the app links time entries to invoices, because many tools track tasks only.
- Salesforce-native PSA. Kantata describes these as adding resource management, project financials, utilization tracking, revenue recognition and portfolio visibility on the same data model as the CRM. Viewpoint Analysis names Certinia as built natively on Salesforce and typically the first shortlist entry for firms centered on Salesforce, and says Salesforce PSA serves overlapping but distinct use cases without explaining the difference. Verify both claims in demos. Ask for a reference. Test both.
- A separate PSA with an integration. Some firms pick a best-of-breed tool outside Salesforce, such as those Viewpoint Analysis lists in its mid-market and specialist groups. The integration then carries accounts, deals and project data in both directions. Ask who builds and maintains it, what happens when either side changes and how errors are caught, since integration failures leave sales and delivery disagreeing about facts. Name an owner for it.
From closed deal to scoped project
A good handoff starts before the deal closes. Sales and delivery should agree on what a standard engagement looks like: phases, deliverables, role mix, rate card and billing milestones. The opportunity record then carries the data a project needs, such as the contract type, the estimated hours by role, the start date and the client’s invoicing rules. When the deal closes, the project is created from that data, and nobody retypes it.
Fixed-fee work deserves special care. The estimate behind the price is the budget, and any difference between the estimate and the delivered hours is margin gained or lost. Record the estimate as hours by role at standard rates, along with the assumptions behind it. When the project starts, the delivery lead can compare actual hours with the plan every week, and the firm can raise a change request when scope moves.
Scope change is the main source of leakage. A client asks for an extra report, a second workshop or a new integration, and the team says yes without a change order. Each favor is small, and together they erode the fee. A system that shows hours against budget as they accrue gives the delivery lead the evidence to say that a request needs a change order, and it gives the partner a way to review the pattern across clients.
Resource planning and staffing
Staffing starts with the pipeline. Kantata’s page says that without pipeline-to-capacity visibility, staffing decisions become reactive. A firm that knows 3 large deals are likely to close next quarter can hire, hold bench or move people accordingly, while a firm that learns from the signed contract is always late. Link the opportunity’s probability and start date to a capacity view, and apply a weighting that reflects your close rates. Review it weekly.
Skills matter as much as headcount. Having 10 consultants available doesn’t help if the project needs a specific certification or industry background. Record skills, certifications and experience in a structure that the staffing process can search. Keep it current, because stale profiles lead to bad matches. Ask people to review their own entries twice a year, and have managers confirm them. Audit the profiles each year. Keep them accurate. Keep notes on changes.
Availability needs a clear rule. Decide how you count leave, holidays, training, internal projects and sales support. The Beancount.io article starts with 2,080 hours a year and notes that leave, holidays, sick days and training must be subtracted, without giving a formula for net availability. Choose your own, document it and apply it consistently, since inconsistent availability makes every utilization number suspect. Publish the rule. Apply it everywhere. Document it.
Staffing is a negotiation, not a calculation. Delivery leads, practice heads and partners compete for the same people, and the system should make trade-offs visible, not decide them. Show who is over-allocated, who is on the bench and which projects are short of skills. Hold a weekly staffing meeting with that view on screen, and record decisions so people can see why changes happened. Record each decision. Share the outcomes.
Utilization: definitions that hold up
Utilization is simple in concept and messy in practice. The Beancount.io article gives billable utilization as total billable hours divided by total available hours, with an example of 30 billable hours in a 40-hour week, or 75 percent. It also defines resource utilization as all productive time, billable and non-billable, divided by available hours, and treats the gap between the 2 as a proxy for overhead load. Choose a definition and stick to it.
| Metric | Formula (per Beancount.io unless noted) | What it tells you |
| Billable utilization | Billable hours divided by available hours | Share of capacity sold to clients |
| Resource utilization | Billable plus non-billable productive hours divided by available hours | Share of capacity doing useful work |
| Available hours | Gross hours minus leave, holidays, sick days and training | The denominator; the article gives no formula |
| Realization | Billed amount versus standard rates | How much of the rate card you actually charge |
| Collection | Cash received versus billed | How much of the billing turns into cash |
| Overhead load | Gap between resource and billable utilization | Time spent on non-billable work |
Utilization alone can mislead. The same article gives an example of a consultant at 90 percent on discounted, written-off or uncollected work, who is busy yet unprofitable. Track utilization together with realization and collection, and review all 3 by role, practice and client. A firm that looks only at the first number may encourage people to log hours that never turn into revenue. Review them together. Track all 3 together.
How to read the benchmarks
The Beancount.io article cites SPI Research’s 19th annual Professional Services Maturity Benchmark, based on a 2025 survey of 509 firms, for an average billable utilization of 66.4 percent, which it calls the lowest in 19 years. It adds that the figure was about 75 percent a decade ago, a claim with no source. The other targets in the article, such as 70 to 80 percent overall and 75 to 80 percent for consultants, have no stated source either.
Benchmarks differ by role and sector, and averages hide the spread. The article says managers and leaders run at 35 to 50 percent, accounting, architecture and advertising firms at about 70 percent, and law firms at 37 percent, citing the Clio Legal Trends Report for the last. A partner who compares the firm’s blended number with a single headline average will draw the wrong conclusion. Compare like with like: same role, same sector and similar size.
Treat targets as ranges and conversations, not rules. The article itself notes that rigid targets can lead to gaming or burnout, which it asserts without evidence. In practice, a target that’s too high pushes people to bill time they didn’t need to spend, and a target that’s too low wastes margin. Set targets by role with input from practice leaders, review them each year and look at realization and client satisfaction alongside them.
Billing models and what the system must support
- Time and materials. Clients pay for hours at agreed rates. The system must capture time by project, task and role, apply the right rate, enforce approvals and produce an invoice that matches the client’s format. Rate changes mid-engagement and different rates by role or location add complexity. Check how the tool handles rate cards, caps and ceiling limits. Test rate changes mid-project in the demo, and ask how approvals and corrections to submitted time are handled.
- Fixed fee. Clients pay a set price for a defined scope. The system must track actual hours against the estimate, show margin as the project progresses and support change orders. Billing may follow milestones or percentages. Because the firm carries the overrun risk, real-time margin visibility matters most here, and it’s the gap that Kantata highlights. Ask for a margin view by project, and check that it updates daily so delivery leads can act before the overrun grows, rather than at month end.
- Retainer. Clients pay a recurring fee for a pool of hours or a defined service. The system must track consumption against the pool, handle rollover rules and flag when a client is under- or over-using. Retainers create hidden overage when delivered hours exceed contracted hours without billing, a problem that shows up as unbilled time. Ask how unused and overused hours are reported to the client, and who approves any carry-over, so the pool never drifts without anyone noticing.
- Milestone and outcome-based. Clients pay when deliverables or results are achieved. The system must link billing events to acceptance, and revenue recognition must follow delivery rather than the deal close date, a point Kantata raises. Viewpoint Analysis advises checking that your primary billing model is supported natively rather than through workarounds, a sound test for any vendor. Ask how acceptance is recorded, who approves it and how a disputed milestone affects recognition, and confirm the answer with finance before signing.
A fixed-fee margin example
Take an invented $120,000 fixed-fee engagement estimated at 600 hours, which implies a $200 effective rate. Suppose cost per hour is $85, also invented. At plan, cost is $51,000 and margin is $69,000, or 57.5 percent. If delivery takes 690 hours, a 15 percent overrun, cost rises to $58,650 and margin falls to $61,350, or about 51.1 percent. The effective rate drops to about $174 an hour. Every input is invented.
| Item (illustrative) | Planned | Actual |
| Fixed fee | $120,000 | $120,000 |
| Hours | 600 | 690 |
| Cost at $85 per hour | $51,000 | $58,650 |
| Margin | $69,000 | $61,350 |
| Margin percentage | 57.5% | About 51.1% |
| Effective rate per hour | $200 | About $174 |
A margin loss of $7,650 on a single project looks small. Across 30 similar engagements, it reaches $229,500. The point is that overruns accumulate unseen unless the system shows hours against budget while the work is under way. Every input is invented, and your costs will differ. Run the same table on your own projects, and add the cost of write-offs and discounts. Run it on your own data. Test your own.
Revenue recognition, invoicing and the finance link
Finance needs delivery data. Kantata’s page says that when delivery and billing data sit outside the core system, finance teams face reconciliation work. Time, expenses, milestones and approvals should flow to invoicing without retyping. Ask your finance lead which reports they build by hand each month, and use that list as a requirements source. Each manual report is a candidate for automation. Start there. Ask finance for the list. Start with their list.
Revenue recognition follows delivery. Kantata names revenue recognition tied to delivery milestones, rather than the deal close date, as a gap in native Salesforce. Viewpoint Analysis highlights ASC 606 and IFRS 15 support for Certinia and lists revenue recognition compliance among selection criteria. Your auditors and finance team should define the rules, and the system should apply them. Verify the claims in a demo using a contract of your own.
Decide where the general ledger lives. The Kantata page cites the 2026 SPI Benchmark for a claim that 76 percent of high-performing professional services organizations use a dedicated PSA, and over 50 percent of those integrate it with their core financial system. Treat the figure as a vendor-cited statistic. The practical question is which system holds the invoice and the ledger entry, and how errors are corrected when they occur.
Plan for multiple entities and currencies. Firms with several legal entities, currencies or tax regimes need the billing and recognition logic to handle them. Kantata lists multi-entity financials among native gaps. Ask vendors how they handle intercompany work, currency conversion and tax, and ask for a reference customer with a similar structure. Complexity discovered after implementation costs far more than complexity discovered in the demo. Test it in a demo.
Selection criteria
- Fit with your main billing model. Viewpoint Analysis lists fit with time and materials, fixed fee, retainer, milestone and outcome-based models, checked natively rather than through workarounds, as the first criterion. Bring examples of your 3 most common contracts to every demo and ask the vendor to configure them live. A tool that handles your contracts without custom work saves time at every invoice. Ask for a live configuration. Insist on a live test.
- Integration with your ecosystem. The same guide asks whether you favor a Salesforce or NetSuite ecosystem or a best-of-breed stack. Map the systems around the tool: CRM, accounting, payroll, HR and time entry. Ask how data moves, how often and who fixes failures. A tool that fits the workflow but breaks the integration map can cost more than a weaker tool that fits the map. Ask for a data map.
- Daily usability. Viewpoint Analysis says usability for consultants and project managers matters because adoption affects results. If consultants find time entry painful, they’ll enter it late and with errors, and every report suffers. Let a few consultants and project managers try the tool on real tasks, and weigh their feedback heavily. Let consultants try time entry on a real week of work and ask them how long it took, what confused them and what they would change before go-live.
- Depth where it counts. Resource management, revenue recognition compliance, portfolio reporting and mobile use are the other criteria the guide lists. Rank them for your firm, and test the top ones in detail. The guide suggests shortlisting 4 or 5 vendors, with sizing guidance such as Kantata for roughly 50 to 500 billable staff, which is the publisher’s view rather than a rule. Test the top ones first. Test those first with real data.
A utilization example with real arithmetic
An illustrative firm has 40 billable consultants, a $200 standard rate and 1,880 net available hours per person after leave, holidays and training, an invented figure. At the 66.4 percent average SPI Research reports, billable hours are about 1,248 per person. At 75 percent, they’re 1,410. The gap is about 162 hours per person, which at $200 is about $32,340 per person, or about $1.29 million across the firm, before any discount or write-off.
| Item (illustrative) | Calculation | Result |
| Net available hours per person | Assumption | 1,880 |
| Billable hours at 66.4% | 1,880 × 0.664 | About 1,248 |
| Billable hours at 75% | 1,880 × 0.75 | 1,410 |
| Difference per person | 1,410 minus 1,248 | About 162 hours |
| Revenue at standard rate per person | 162 × $200 | About $32,340 |
| Firm total for 40 consultants | 40 × $32,340 | About $1.29 million |
| With 90% realization | $1.29 million × 0.90 | About $1.16 million |
The Beancount.io article offers a smaller check, in which 1 more billable hour a week across 48 weeks at $200 is $9,600 per person, or $96,000 for 10 people. Scale it to 40 consultants and it’s $384,000. These are standard-rate values and ignore realization, so the real figures are lower. They show why small utilization gains matter, and why firms invest in tools that make capacity visible. Verify the math.
Implementation and adoption
Start with the data model. Decide how accounts, opportunities, projects, tasks, resources, time entries and invoices relate, and who owns each record. Write it as a diagram that sales, delivery and finance can read. Disagreements surface early when people can see the same picture, and the diagram becomes the reference for configuration, reports and integrations. Revisit it when the business adds a new service line or billing model. Keep it current as the firm changes, and store it where all 3 groups can find it.
Plan for change management as much as configuration. Consultants will resist time entry that feels like surveillance, project managers will resist forecasts that expose their risk and partners will resist rules that limit their discretion. Explain why each data point is needed and how it’ll be used. Share the first reports with the people who generate the data, and fix their complaints quickly so they see that the system serves them.
Roll out in phases. Begin with 1 practice or service line, a limited set of billing models and a small group of users. Run the new process alongside the old process for 1 or 2 billing cycles, compare results and fix gaps. Then extend to more teams. A phased rollout reduces risk and builds a group of experienced users who help the next wave, which matters more than a perfect configuration on day 1.
Mistakes to avoid
- Buying a tool before defining the questions. Firms often start with a demo and end up with a product that answers questions nobody asked. List the decisions leaders make each week, the data they need and who needs it. Use that list to score vendors, and discard features that don’t serve it. A shorter requirements list produces a faster, cheaper project. Write the questions first. Keep the list short and ranked.
- Treating utilization as a single target. Pushing everyone toward the same number ignores role differences, and the Beancount.io article’s own ranges show managers well below consultants. Set targets by role, and review them against realization, margin and client outcomes. A target that rewards logged hours without regard to revenue leads to gaming and weakens trust in the data. Review it each year with practice heads, and adjust targets when roles or services change.
- Ignoring the handoff from sales. If scope, price assumptions and billing terms don’t pass cleanly from the opportunity to the project, delivery works from a different picture than the scope the client bought. Define the handoff, include a short review meeting for large deals and capture the estimate in a form the project can use. The cost of fixing a bad handoff grows with every week. Test it. Review it quarterly.
- Skipping finance. A system that delivery loves but finance can’t use creates reconciliation work and delayed invoices. Involve the finance lead from the start, and test billing, recognition and reporting with real contracts before go-live. Run a parallel close for the first cycle, and fix discrepancies before turning off the old process. Ask the finance lead which reports they build by hand each month, and make those reports the first test cases for the new system, run in parallel for a full cycle.
A 90-day starting plan
A firm that wants to start without committing to a large project can use the first 90 days to build a baseline. In the first 30 days, define utilization, availability and realization, pull 12 months of time and billing data and compute the figures by role. On days 31 to 60, map the handoff from opportunity to project and list the questions leaders can’t answer. On days 61 to 90, run demos against those questions.
By day 90, the firm should have 4 things: a written definition of each metric, a baseline for the last year, a ranked list of gaps and a shortlist of 2 or 3 options with costs. Viewpoint Analysis suggests shortlisting 4 or 5 vendors and describes selection timelines of weeks, so a firm that has done the groundwork can move quickly. Leaders then decide with evidence instead of impressions. Present them to leadership.
Budget for the whole cost, beyond licenses. Include implementation, integration, data migration, training, ongoing administration and the time of your own people. The sources give no prices, so request written quotes for your user count and requirements, and compare them on a 3-year basis. Ask each vendor for references from firms of similar size and billing mix, and call at least 2 of them. Ask for 3-year totals. Ask for written quotes.
FAQs
Can Salesforce handle project management for a professional services firm?
For light needs, yes. Test it more. Kantata’s page says tasks, activities, Flow and AppExchange tools support basic tracking, and that the ceiling becomes clear quickly for firms with several concurrent engagements. It also says Salesforce doesn’t provide skills, billable utilization or capacity visibility on its own. Kantata sells a competing tool, so test the claim against your requirements. Test it. Ask vendors to demonstrate. Test it with real data.
What’s billable utilization and how do I calculate it?
Beancount.io gives billable utilization as total billable hours divided by total available hours. For example, 30 billable hours in a 40-hour week is 75 percent. Available hours should exclude leave, holidays, sick days and training, and the article gives no formula for net availability. Define yours, apply it consistently and track it with realization and collection. Track it by role and month. Review it monthly with practice leads. Check it often.
What’s a good utilization rate?
It depends on the role and sector. The Beancount.io article cites SPI Research for a 66.4 percent average in 2025, and gives unsourced targets of 70 to 80 percent overall, 75 to 80 percent for consultants and 35 to 50 percent for managers. Set targets by role, review them each year and look at realization and margin alongside them. Compare like with like. Revisit targets each year with practice heads. Compare ranges, not single numbers.
Do I need a PSA tool on top of Salesforce?
It depends on your size and billing complexity. Pilot it first too. A firm with a few projects and a simple billing model may manage with native tools. Firms with many concurrent engagements, fixed fees or milestone billing usually need resource and financial visibility. Kantata cites a figure that 76 percent of high performers use a dedicated PSA, which is a vendor-cited statistic. Ask for a demo. Ask vendors for references. Pilot first.
What should a PSA support for billing?
It should support your primary billing model natively, whether time and materials, fixed fee, retainer, milestone or outcome-based. Viewpoint Analysis lists this as a selection criterion, along with resource management depth, revenue recognition compliance and portfolio reporting. Test each model in a demo using your own contracts and rate cards. Ask for a live demo of each model, using a sample contract of your own, and have finance sit in on it.
What does it cost?
The sources give no prices for any tool. Viewpoint Analysis says it provides no price ranges or licence costs for any vendor. Request written quotes for your user count, and add implementation, integration, training and administration to the comparison. Ask for a 3-year total, and check how prices change when you add users or entities. Get quotes in writing and compare them over 3 years. Ask for itemized quotes. Compare totals.
How do I handle revenue recognition?
Kantata names revenue recognition tied to delivery milestones, rather than the deal close date, as a gap in native Salesforce, and Viewpoint Analysis highlights ASC 606 and IFRS 15 support in Certinia. Your auditors and finance team should define the rules, and the tool should apply them. Verify in a demo using a real contract. Ask finance to attend the demo and test a milestone case from your own contracts.
How long does an implementation take?
The sources don’t give an implementation timeline. Viewpoint Analysis describes selection timelines of weeks, with no delivery timelines. Plan from your own scope, number of billing models, entities and integrations. A phased rollout starting with 1 practice and 1 or 2 billing cycles of parallel running reduces risk, and a partner can estimate the effort after a short discovery. Ask vendors for references. Ask for a plan. Plan in phases.
