The Add-On Sales Gap: What Your Techs Find and What Gets Quoted

Your Best Sales Team Doesn't Know They're Selling

Your techs see the equipment while they are on site: aging components, compressor strain, corroded piping and damaged ductwork. That access creates a chance to document a need. It does not mean every visit should produce an add-on sale.

A finding can stop at the technician’s notes, never becoming a quote or a scheduled repair. That handoff is worth inspecting before increasing visit volume.

An earlier Level analysis covered 54,000+ service agreements across 240 contractor records and $412M in maintenance revenue. Its trade and book-size tables below are preserved as historical context, with their interpretation limits. The published Level Index uses a separately defined 386-company population. To act on either, connect visit findings to approved quotes, completed work and its actual gross profit.

Pull-Through: The Metric That Separates Growth from Survival

Pull-through revenue is additional repair, replacement and project revenue from service-agreement customers. The Level Index divides it by annual service-agreement revenue. It is a revenue ratio, not the share of visits or customers that buy, and can include relationship project work that no maintenance visit uncovered. Pair it with the actual visit-to-finding-to-quote funnel; one cannot stand in for the other.

Why this matters for lifetime value: compare cumulative gross profit, service costs and retention for matched SA and non-SA customer cohorts. Do not assume agreement customers are worth a fixed multiple. The Level Index’s separately defined service-agreement gross margin ranges from 20.5% at P25 to 53.5% at P75 (259 eligible companies with at least $10K in SA revenue). An agreement can create repeat access while still losing money on covered work.

Do not confuse annual pull-through with lifetime customer value. The published ratio is 8.7% at the median and 29.6% at P75 across 386 eligible companies. Lifetime economics also depend on renewals, churn, service cost, subsequent projects and the observation window. Those facts are not supplied by the annual ratio.

Fictional illustration on a $500K annual agreement book: at 8.7% pull-through you add roughly $43.5K/year of additional revenue. At 29.6% you add about $148K/year, a ~$104.5K annual revenue gap from the same customer base. That gap is revenue, not profit. At an assumed 40% gross margin on the add-on work, it is about $41.8K of gross profit. It is also not all recoverable: some agreement customers have no further work to buy, and part of the P75 group's figure is project work tied to the relationship.

A note on measurement: the published Level Index divides additional repair, replacement and project revenue from agreement customers by annual service-agreement revenue. That produces a median of 8.7% across 386 companies. The earlier 240-contractor cut used maintenance revenue as the denominator and reported a similar median (8.9%) and 75th percentile (29.5%). That cut's sample and denominator were not re-derived for this update, so the table below uses the published distribution.

Level Index annual service-agreement pull-through (n=386 eligible companies, cross-trade, not separately measured trade subsets):

*P90 caveat: The 93.4% figure includes contractors where relationship-driven project work (chillers, retrofits, tenant improvements) is attributed to the Service Agreement customer relationship. This inflates pull-through beyond what pure "tech finds issue on visit" upselling produces. P75 (29.6%) is a comparison point, not a proven attainable target for your agreement book.

The 75th percentile is about 3.4x the median (29.6% versus 8.7%). Companies at P75 generate additional revenue equal to roughly 30% of annual agreement revenue. The distribution does not say why. Follow-up process, equipment age, customer mix and how project work is attributed are all candidate explanations to check in your own records.

The earlier 240-company cut reported $1.7M average maintenance revenue and $564K average additional revenue. Dividing the two averages gives about 33%, but that ratio need not equal a median company’s pull-through. Large accounts and different weighting can move it; the original calculation’s aggregation rules have not been re-derived here. Compare the separately published median and P75 before choosing a goal.

We covered the pull-through mechanics in our account penetration analysis, but the tables here help separate revenue attribution, agreement delivery and the quote handoff.

Add-On Performance by Trade

In the earlier 240-contractor cut, add-on rates vary by trade. Samples are small for several trades, and this cut is not a separately measured trade benchmark.

Correction note: an earlier version of this table also showed average pull-through revenue by trade. Those dollar figures did not reconcile with the percentage column. For electrical, $68,756 divided by $6,254 is about 11x, not 25.3%. The underlying aggregation has not been reproduced, so we cannot establish whether each column is a mean per agreement, a mean per company or a median of company ratios. The dollar column is removed until it can be re-derived. Read the remaining columns as historical directional context, not reconciled current trade benchmarks. The displayed agreement counts sum to 53,357; the original broader 54,000+ headline has not been reconciled to these rows. The 174 electrical and 276 plumbing agreements are thin samples.

Three things jump out:

Mechanical’s historical cut is 45.9%. Inspect whether the numerator includes large chiller, retrofit or project work from agreement accounts. That revenue belongs in relationship economics, but it does not prove that a technician sold it during a maintenance visit.

HVAC sits at 29.7% in this cut. With 11,244 SAs, HVAC has the largest named-trade sample. The supplied aggregate table does not establish follow-up behavior or show that follow-up causes the result. To test that, measure findings, quotes issued and quotes won in your own records.

Plumbing’s historical cut is 5.5%, with 276 agreements. Inspect the covered inspection scope and the types of additional work a visit can identify. This thin agreement sample cannot establish that all plumbing firms underinvest in training or that their opportunity is greater than HVAC’s.

The Service Agreement Attach Rate Problem

Before your techs can upsell during maintenance visits, you need customers on maintenance agreements in the first place. The Service Agreement (SA) attach rate, what percentage of your service customers also have an SA, defines the eligible relationship base before a maintenance-visit add-on conversation starts.

The ranges below are illustrative planning assumptions. We could not trace their source or population for this update, and they are not Level Index measurements:

If your HVAC attach rate is near 7-8%, moving from 8% to 32% would multiply the rate by four. Across the entire displayed 7-8% to 30-40% ranges, the multiple is about 3.75x to 5.71x; those assumed endpoints are not one measured improvement. Measure your own rate first, using the same 12 months for both counts: active SA customers divided by unique service customers. Then set a target.

A separate Level cut of 1,833 companies compared companies with and without SA programs. It is not the 240-contractor or n=386 sample and was not re-derived for this update:

Companies with SA programs averaged about 2x the job volume (1,178 vs 574) and about 4.8x the revenue ($6.3M vs $1.3M). This is a comparison of averages, not a controlled test. The SA group has higher average revenue, but group averages alone do not establish how adoption probability changes with company size or show that SAs cause growth. A feedback loop, in which SAs create repeat access and repeat access creates add-on work, is a hypothesis to test with your own cohort data. We explored the valuation side in our analysis of maintenance vs install revenue impact on valuation.

What Service Agreement Book Size Tells You About Add-On Maturity

Bigger SA books should mean better pull-through, right? The earlier cut used a different frame from the 240-contractor and n=386 samples: 549 companies with book-size data, not re-derived for this update. In that cut, the pattern is not linear:

The smallest books show the highest average pull-through in this historical cut. With a small agreement-revenue denominator, one relationship project can move the ratio sharply. Check project attribution before interpreting the figure as better visit execution. The table does not identify the cause.

The 50-200 group averages 29.5%, and the 200-500 group 19.0%. These are different groups of companies, not the same firms observed as they grow. Check findings per visit, quotes created, days to follow-up and customer mix in your own records before diagnosing a maturity problem.

The 500+ group has 25 companies, averages 51.8% pull-through and reports 72.3% visit completion. The cut does not establish which systems those companies use or whether the system caused either figure. Visit completion and attributed project revenue can change independently.

If your SA book is between 50-500 agreements and your pull-through is under 20%, start by measuring the conversion steps (visits completed, findings, quotes, wins) before adding more agreements.

The Upsell Math: $150/Call vs $700/Call

The ticket bands below are illustrative planning inputs, not measured training outcomes. Annual revenue is recalculated using the same assumed 1,000 completed calls in both columns; a higher ticket can also take more time:

Fictional example: moving from a $200 average ticket to a $600 average ticket adds $400 of revenue per call. At an assumed 1,000 calls per tech per year, that is $400K of additional revenue, not profit. Add-on work takes extra labor time and materials. At an assumed 40% gross margin it is about $160K of gross profit, and calls per year may fall if each call runs longer. Capacity counterexample: a fictional $300 ticket at 1,000 calls and 50% gross margin produces $150K gross profit. A $500 ticket at only 750 calls and 40% margin also produces $150K, before training cost. Revenue rises from $300K to $375K while gross profit does not. Track completed calls, direct costs and total gross profit together.

Illustrative planning ranges (unsourced, not Level Index measurements):

A service agreement can create repeat access, but its economics depend on what you cover, how often you return and what customers actually need. Track offers on eligible calls and gross profit per retained customer. See the renewal-rate analysis, keeping cancellation events, nonrenewals and customer retention as separate definitions.

Replacement Conversion: The Highest-Dollar Add-On

When a tech identifies equipment that needs replacement, the conversion math looks different from repair close rates. These bands are illustrative planning ranges, not measured benchmarks:

For context, the Level Index reports 73.9% median conversion on decided quotes across 794 eligible companies. That is won / (won + lost), not a replacement-only close rate. Its recorded winning quotes have a two-day median sent-to-accepted interval; this does not show that a two-day follow-up causes a win. For replacement work, measure quote approval, customer decision timing and gross profit on the work won.

Check your own repair trend: revenue and gross profit per completed repair, by year and comparable call type. Separate price changes, materials, extra labor time and job mix before concluding that training increased revenue per call.

The Revenue Expansion Flywheel

Use this four-step operating model to test the handoff from a service visit to relationship economics:

Stage 1: Service call → Record whether an eligible customer was offered an agreement and accepted it

Stage 2: Maintenance visit → Identify work, generate recommendations (Level Index pull-through: 8.7% median, 29.6% P75)

Stage 3: Recommendations → Convert findings to approved quotes, then record decisions, work completed and gross profit

Stage 4: Expanded relationship → More renewals, cross-sell and referrals → Compare cumulative customer gross profit and service cost

Fictional example: a $200 annual agreement account buys $60 of additional work in that year. Pull-through is $60 / $200 = 30%. The customer might buy nothing in the next year or approve a large project; neither outcome tells you how many maintenance visits converted. Track those visits separately.

For lifetime economics, measure cumulative gross profit by customer cohort, subtract incremental acquisition and support costs, and state the number of years observed. No fixed HVAC or plumbing lifetime-value amount follows from this agreement ratio.

How to Close the Gap

Practical steps to test, framed as operator checks rather than proven causes:

1. Measure pull-through by tech, by trade, by SA. You can't improve what you don't track. Check whether your records can tell you which techs generate recommendations, which SAs produce add-on work, and which trade departments convert at what rate. Our financial dashboard guide covers how to set up this visibility.

2. Build a recommendation workflow, not just a checklist. The benchmark distribution does not say whether skill or process explains the gap between 8.7% and 29.6%, but process is the part you can inspect quickly. Check whether you have: (a) standardized condition reporting during every visit, (b) a quote created from each recorded finding, and (c) a named person who follows up within 48 hours. The quote timing data in our follow-up analysis is an association, not proof that a 48-hour cadence creates wins. Test whether earlier handoff changes qualified quote decisions in your own cohort.

3. Train techs to recommend, not sell. One workable model is for techs to document everything they find and for the office to convert findings into quotes and follow-up. The tech's job is identification and trust-building. The office's job is the close.

4. Set an SA attach rate target. Set an attach-rate goal from your own eligible customer base and service capacity. For a fictional 1,000-customer base, moving from 80 to 150 active agreements changes attach rate from 8% to 15%. It also creates covered work to deliver. Check cancellation, visit capacity and agreement gross profit before treating those 70 agreements as a gain. Your customer profitability analysis should distinguish between SA and non-SA accounts to prove this internally.

5. Segment your SA book by pull-through potential. Not all SAs are created equal. A $2K/year agreement on an older chiller plant and a $500/year agreement on a new RTU expose different equipment and service obligations. Inspect age, condition, covered scope and replacement timing; the agreement price does not establish a multiple of future work. Prioritize high-potential SAs for your best techs and most thorough inspections.

The Bottom Line

The benchmark gap is large, but not all of it is recoverable by any one company. The comparison points:

• Pull-through: 8.7% median vs 29.6% at P75 (Level Index, n=386; P90 of 93.4% includes relationship-driven project work)
• SA attach: active agreement customers / eligible unique service customers over a defined window
• Tech ticket: revenue and gross profit per completed comparable call, with labor hours and materials included
• Replacement conversion: wins / decided replacement quotes, with pending quotes tracked separately

Moving from the Level Index median to P75 on pull-through, 8.7% to 29.6%, would mean about 3.4x the add-on revenue from the same customer base. In the fictional $500K book above, that is about $43.5K vs $148K of annual add-on revenue, a ~$104.5K/year revenue difference. It is revenue, not profit, and the extra work needs labor hours and materials. Judge it on gross profit (about $41.8K at an assumed 40% margin) and on how much of the P75 figure reflects project work your customers may not have.

A useful finding reaches a customer decision only if someone owns the handoff. Measure that step before assuming a larger ticket is available.

Frequently Asked Questions

What's a good pull-through rate for contractors? In the Level Index, the median is 8.7% and the 75th percentile is 29.6%. That metric covers 386 eligible companies and divides additional revenue from agreement customers by annual agreement revenue. Above 29.6% puts you above P75. P90 is 93.4%, but it includes relationship project work. The distribution does not identify the driver, so check your own conversion steps: findings per visit, quotes created and quotes won.

How do I calculate my SA attach rate? Divide the number of customers with active service agreements by your total unique service customers. The 7-8% HVAC and 30-40% stretch ranges used on this page are unsourced planning assumptions. Use the same 12-month window for numerator and denominator, track the rate monthly and set targets from your own baseline.

What's the ROI of training techs to upsell? Measure training on matched call types, not an unmatched comparison between technicians. Fictional example: a $100 ticket lift at an assumed 40% incremental gross margin adds $40 of gross profit per call. A $4,000 training cost needs 100 such calls, or five weeks at 20 per week. That is a conditional calculation before lost call capacity or extra overhead; with no incremental contribution, there is no payback.

Why is plumbing pull-through so much lower than HVAC or mechanical? The earlier cut reports plumbing at 5.5%, mechanical at 45.9% and HVAC at 29.7%; plumbing has 276 agreements. Those values do not establish the cause or rank the total dollar opportunity for each trade. Inspect the covered work, equipment, quote handoff and project attribution in your own accounts. A small denominator can amplify a percentage without creating more gross profit.

How do I improve from the 50-200 SA "trough" where pull-through drops? In the earlier cut, average pull-through was lower at 50-200 SAs (29.5%) and 200-500 SAs (19.0%) than in the smallest and largest books. The cut does not show why. Practical checks: designate a follow-up person (not the tech), measure days from finding to quote, and track pull-through by tech and by account monthly.

Should I set add-on revenue targets for individual techs? Yes, but frame it as recommendation volume, not sales dollars. Avoid incentives that reward unnecessary work or recommendations without evidence. Review recommendation quality and customer consent alongside volume; documenting findings alone does not establish that a repair is needed. Track the number of recommendations per visit, then measure what percentage the office converts. This separates the identification responsibility (tech) from the closing responsibility (office).

Level helps contractors measure and improve their add-on sales performance. Book a free profitability audit and we'll compare your pull-through against the published Level Index distribution and help you measure your attach rate and tech-level ticket trends from your own records.