Competitor Sales Capacity Teardown: From Rep Count to Bookings Range

Leadership has named a competitor and asked how fast it plans to grow. This guide is for the strategy, research or talent-intelligence analyst who has to answer with two things: a count of the competitor's real quota-carrying reps by segment, and a bookings range that holds up under questioning. I follow one anonymized competitor through every step. I use the real counts from Refolk's index, show the decision points where the answer can double, and end with a worksheet you can reuse on your own case.

Why does the one-line rule of thumb overshoot?

The best-known shortcut counts a competitor's reps, multiplies by a flat per-rep figure and cuts about 40% at larger companies. It overshoots for two reasons. Most profiles in a raw "sales" count do not carry a quota, and the flat figure blurs the difference between quota and what reps actually book.

Jason Lemkin's SaaStr version says "Multiply by $500,000 ($250,000 if SMB)", then cut the result by about 40% "where the # of reps seems to be > 100". It is designed to give an answer in about 60 seconds, and it works as a first rough bound. As a number to put in front of leadership, it has three gaps:

• Role mix. A "sales" title search returns SDRs, sales engineers, customer success managers, sales leaders and unrelated titles, not just the people who close new business.
• Freshness. Some profiles still list jobs the person has left, or employers that no longer exist under that name.
• Ramp. A rep hired three months ago does not yet produce a full rep's bookings, and how far short they fall depends on the segment they sell into.

Sources disagree on how to correct for these. SaaStr makes one flat cut at the end. ORM's capacity guidance adds a separate correction for each of attainment, ramp and attrition. I follow ORM's approach. Each correction is stated separately with its basis, so someone who disagrees can challenge one assumption without throwing out the whole estimate. The flat rule stays in as a sanity check.

Terms used throughout:

• Quota carrier: a rep with an individual new-bookings target. In this model, that means account executives (AEs).
• Quota: the target a rep is given.
• Yield: what a rep actually books, or what a plan expects them to book once misses are allowed for.
• Attainment factor: what the company expects its reps to deliver, divided by the total quota it assigns them. Revcast's example: a $1,000,000 street quota against an $800,000 company target gives an 80% attainment factor.
• Ramp: how many months a new rep takes to reach full productivity. Lative counts a rep as ramped at 80% of quota.

Every estimate in this guide comes from one practitioner formula: ramped productive rep months × quota per rep month × attainment = bookings.

The worked example: Company A's raw pull

The raw pull is one broad title search for each alias and each role family. Log every count next to the exact query that produced it, including searches that return zero. The worked example is a US B2B sales-software company I call Company A, pulled from Refolk's index of professional profiles.

Before you search, fix the entity: list the competitor's legal names, acquired brands and regional entities. This matters more than it seems, because people keep old employer names on their profiles for a long time (see the freshness section). Company A had a short alias list, so one name covered it.

I searched two role families in the Computer Software industry tag: Account Executive titles and SDR/BDR titles. I ran the AE search in the US and the UK. Here is what Refolk's index returned:

The two slices differ in how clean their titles are. Of the 22 AE profiles, 20 (91%) carry some version of the AE title. In the SDR sample, the most common title covers only 17 of 25 (68%). Both slices are also dominated by one employer, and in both cases it is Company A.

The first wrong turn: reading the UK zero as a fact

The UK AE search returned no profiles. It is easy to write "no UK sales presence" in the deck and move on.

I don't have a re-run count to publish for this slice. So the UK stays as an open line on the worksheet, not as a zero in the estimate. On your own case, run the unfiltered search before you go to step 3.

Separating quota carriers from everyone titled "sales"

In this model only account executives carry a new-bookings quota, so remove every other role before any dollar figures go in. In Refolk's index this step changes the answer by a multiple, because SDRs outnumber AEs.

Give every profile exactly one label. For each label, know what it tells you and how it can mislead:

• AE: tells you the person closes deals. It misleads when the title is inflated or someone senior uses it loosely. One of the 22 AE profiles was a CEO.
• SDR/BDR: tells you the person builds pipeline and has no bookings quota. It misleads when the title string is out of date. One profile in the SDR sample was actually an Account Executive.
• SE (sales engineer): supports deals but carries no quota of their own in this model.
• AM/CSM: handles expansion and retention. Leave them out of new-bookings capacity unless the competitor's own job posts say they carry a new-logo quota.
• Leader: managers and directors. Leave them out.
• Noise: everything else. The SDR sample included a Field Technician and an Independent Business Owner.

Do not work backwards from published role ratios to an AE count. The Bridge Group's SDR:AE ratio is well sourced. A 2:1 AE:SE ratio for enterprise appears in one secondary compilation, but I could not verify it. I found no published CSM:AE ratio at all.

The ratio fork

The raw ratio is 51 SDR/BDR profiles to 22 AEs, or 2.3:1. The Bridge Group's SDR study covered 351 B2B companies, mostly North American SaaS with a median deal size of $50K. It found 1 SDR for every 2.4 AEs, a ratio that has held across editions. The most common model is 1 SDR to 2 AEs (31% of companies), and smaller companies run more SDRs per AE. By that benchmark, the slice is 5.5 times more SDR-heavy than normal.

A common wrong turn is to read the benchmark upside down, as two SDRs per AE. Read that way, 2.3:1 looks normal and the analyst moves on.

The right next move is to remove the top employer and recompute. Company A accounts for 21 of the 25 sampled SDRs (84%). Without Company A, 16 AEs remain. Only 4 of the 25 sampled SDRs work elsewhere, which scales to roughly 8 of the 51. That gives about 0.5:1, close to Bridge's 0.42:1. So the wider market looks normal, and Company A is the outlier. Scaled the same way, Company A has roughly 43 SDRs against 6 AEs, about seven per AE.

There are two possible explanations:

1. Company A runs an unusually SDR-heavy sales motion.
2. The AE search misses some of Company A's closers, because they use titles outside the AE family or sit outside the industry tag.

The slice cannot tell these apart. My decision was to carry 6 forward as a floor, label it as a floor, and add "re-run AE search with title variants and no industry filter" to the worksheet.

What the flat rule would have said

Company A has about 49 sales-titled profiles: 6 AEs plus roughly 43 SDRs. That is under 100, so SaaStr's 40% cut does not apply. 49 × $500K = $24.5M. Keep that number in mind; the step-by-step method lands far lower.

How big should the freshness haircut be?

Base the haircut on measured B2B data decay, compounded over the months people take to update their profiles. For a 6 to 12 month lag that gives 12% to 31%. Before applying it, remove ghost employers by searching for them. A flat 40% is an assumption standing in for a rate you can measure.

Freshness involves two separate removals.

Ghost employers, removed by string

A ghost employer is a company that has been acquired or shut down but still appears as someone's current employer. Two of the 22 AE profiles showed this in the employer name itself: "Solvvy (now part of Zoom)" and "Host Compliance - Now a part of Granicus". Search employer names for "now part of", "now a part of" and "acquired". Removing those two profiles and the CEO leaves 19 in the market view. None of the three removals is needed for Company A's count of 6, which carries forward. The decay haircut covers whatever cannot be checked one profile at a time. This is also why step 1 matters: your competitor's acquired brands will show up the same way, and you need to decide whether to count them.

Decay, applied as a percentage

MarketingSherpa's 2.1% a month is the actual measurement; HubSpot annualizes it to about 22.5%. Gartner puts global business-data decay at around 3% a month. Nobody has published a measured stale share for searches of current sales employees, so these rates stand in for one. Say so in your write-up. The title-noise row is a different correction. If you read every profile in step 3, you have already removed that noise. If you only read a sample, apply the sample's noise rate to the profiles you did not read.

Applied to Company A's 6 AEs:

• Low: 30.6% cut (Gartner rate, 12-month lag) leaves 4.16.
• Base: 22.5% cut (MarketingSherpa rate, 12-month lag) leaves 4.65.
• High: 12% cut (MarketingSherpa rate, 6-month lag) leaves 5.28.

SaaStr's 40% is not just a larger number. By its own terms, it does not apply to a six-person team.

The same research is useful elsewhere. People who edited their profiles were 64% more likely to change employers around the same time. Treat an AE profile edited in the last few months as a risk of short tenure: flag it, but do not use the edit rate as a haircut.

Assigning segments and dating ramp

Give each AE a segment based on the words in their title and territory description, and check that against the competitor's own job posts. Then use start dates to convert recent hires into ramped-rep equivalents. This matters because a first-year enterprise AE delivers only about half a year of capacity.

Segment titles mean different things at different companies

Some title strings are documented in public job pages:

• GitLab's AE job family uses "Commercial Account Executive - Mid-Market" and "Commercial Account Executive - Named".
• commercetools says its Named AE works "within a small set of high-value, named enterprise accounts in the South DACH region". There, "Named" means enterprise.
• Square's Territory AE "sells Square's ecosystem to local businesses", which is SMB.
• A cybersecurity firm's Senior Territory Account Executive "will drive new business acquisition and customer expansion in the mid-market segment".

"Territory" means SMB at one company and mid-market at another. Nobody has published how often companies reuse the same title across segments. To check for your competitor, read its current job posts before you map any title to a segment. If a profile does not match, put it in the unassigned bucket rather than forcing it into a segment.

Two more clues help set dollar figures for each segment. If the competitor's job posts show on-target earnings (OTE), use the median quota-to-OTE ratio of 4.2x: a $200,000 OTE typically means a quota near $840,000. As a check, Bridge's median OTE of $190K × 4.2 comes to about $800K, which matches Bridge's median quota. Bridge also found that quota differs by nearly 2.5 times between a seller with ACV under $25K and one with ACV of $250K or more. Of all your inputs, the segment choice moves the per-rep figure the most.

For Company A, this walk-through does not resolve the segment, so all six AEs are unassigned. I handle that by bounding the estimate: the low case assumes SMB, the base case mid-market and the high case enterprise. That widens the range instead of hiding the uncertainty.

Pulling the AE and SDR lists for a competitor and its aliases is the slowest part of this work by hand. Refolk lets you ask for the role, industry and country in plain English and returns the matching profiles, so each pull in step 2 takes one sentence instead of a separate search string for every source.

Ramp: turning start dates into ramped reps

Lative's ramp ranges, measured to 80% of quota, are similar: 3-4 months for SMB, 5-7 for mid-market and 9-12 for enterprise. The last column is my own calculation. It assumes a linear ramp and uses the midpoint of ORM's range: for enterprise, 12 months minus 5.25 lost months gives 6.75/12, or 56%. The only published month-by-month curve is an illustration (0% in month 1, 25% in month 2, 50% in months 3-4, 75% in months 5-6, 100% from month 7), and it works out to about 73% in the first year. Nobody has published ramp curves measured from real data for each segment.

To apply this, take each AE whose start date falls inside the ramp window and count them as a fraction of a rep, based on how many productive months they will have in the next year. A hiring burst cuts first-year capacity far more in enterprise than in SMB, so read start dates most carefully on enterprise teams. Company A's start dates are not resolved here, so the low case treats the whole team as first-year and the base and high cases treat everyone as fully ramped.

From ramped reps to a low, base and high bookings range

Multiply ramped reps by a per-rep dollar figure. Apply an attainment factor only when that figure is a quota. Present three cases, and for each one name the query, haircut and benchmark behind it.

For the quota route, the benchmark is The Bridge Group's AE metrics and compensation report. It surveyed leaders at more than 170 B2B SaaS companies and found a median AE quota of $800K in ACV, up from $740K in the previous edition. In the same report, 51% of reps hit quota, down from 66% in the previous edition and from 74% in the earliest edition on record. Before you rely on these figures, check that your competitor resembles the sample: 87% of the companies are headquartered in North America, with median annual revenue of $24m and a median ACV of $47K.

That 51% is the most commonly misused number in this work. It is the share of reps who hit quota, not the average share of quota that reps deliver, so quota × 0.51 understates bookings. Use an attainment factor instead. Revcast's 80% is a worked illustration, not a measured benchmark. Still, $800K × 80% = $640K, close to SaaStr's $600K mid-market yield, so the quota route and the yield route agree. Quotas have risen about 2% a year while the share of reps at quota has fallen, which means quota × headcount overstates bookings more now than it once did.

Company A's numbers:

To check the base case by the quota route: 4.65 × $640K comes to about $3.0M, which falls inside the range. The flat rule's $24.5M is nearly 9 times the base case. For the flat rule to be right, Company A would need about 41 fully ramped mid-market AEs, and the slice shows 6. The title-variant re-run could raise the AE count, but it would have to find several times more AEs to close that gap. The range is the implied annual new-bookings capacity of the AE team you can see, with every assumption written down. It is not a revenue figure.

Competitor:
Alias list:
Query log (exact text :: count :: date run):
  AE titles ::
  SDR/BDR titles ::
  Same queries, no industry filter ::
Role split (AE / SDR-BDR / SE / AM-CSM / leader / noise):
Ghost employers removed (employer string :: count):
Freshness haircut (percent :: basis :: lag assumed):
Segment split (SMB / mid-market / enterprise / unassigned):
Ramp treatment (start dates seen? first-year factor used):
Per-rep dollar input (figure :: QUOTA or YIELD :: source):
Attainment factor (only if input is QUOTA :: measured or illustrative):
Bookings range (low / base / high):
SDR:AE ratio (with top employer / without):
Independent estimate (method :: figure):
Gap and written explanation:
Open items (e.g. UK re-run, title variants):

The procedure, step by step

This is the full method in order. Steps 1, 2 and 7 come from documented practitioner formulas. The checks inside the other steps come from the failure modes in the next section. The role split takes the longest, because you read every profile or a stated sample.

How this goes wrong: failure modes and false positives

Most bad capacity estimates fail in one of eight ways, and four of them can double the answer or worse without anyone noticing. Each one below has a symptom you can spot and a check you can run.

1. Mixing up attainment measures. What it looks like: quota × 0.51. Why it misleads: 51% is the share of reps who hit quota, not average attainment, so bookings come out too low. Check: use an attainment factor (company target divided by total quota) and say whether it is measured or illustrative.
2. Discounting twice. What it looks like: SaaStr's $500K multiplied by an attainment rate. Why it misleads: $500K is planned bookings per rep, which already allows for misses, while Bridge's $800K is quota. Check: label every dollar input QUOTA or YIELD on the worksheet.
3. Reading the ratio upside down. What it looks like: treating 2 SDRs per AE as normal. Why it misleads: the benchmark is 1 SDR to 2.4 AEs, so the upside-down version makes an SDR-heavy count look normal and a flat count look like mostly AEs. Check: in Refolk's index slice, SDRs outnumber AEs 2.3:1, which means a flat "sales" count is mostly people without a quota.
4. Treating too few results as a real answer. What it looks like: the UK search returns 0 and the deck says "no UK team". Why it misleads: the gap is in the industry tag, not the market. Check: re-run without the industry filter before concluding anything.
5. Ghost employers. What it looks like: profiles listing an acquired company as the current employer, which was 2 of 22 in the AE slice. Why it misleads: those reps either sell under the acquirer's name now or have left. Check: search employer strings for "now part of" and "acquired", and decide whether acquired brands are on the alias list.
6. Assuming a title means the same segment everywhere. What it looks like: every "Territory AE" mapped to SMB. Why it misleads: "Territory" means SMB at Square and mid-market at a cybersecurity firm. Check: read the territory description and the competitor's own job posts before mapping, and keep an unassigned bucket.
7. Using the profile-edit rate as a haircut. What it looks like: a 19.7% cut justified by the NBER paper. Why it misleads: that figure is about edits to past jobs, not whether people still hold their current job. Check: base the haircut on the decay rate plus the count of non-target titles you actually found.
8. One employer dominating the results. What it looks like: a market-level ratio you then apply to a single competitor. Why it misleads: Company A is 84% of the SDR slice, so the "market" ratio is mostly Company A's ratio. Check: remove the top employer and recompute. Here the rest of the market fell to about 0.5:1, close to Bridge's benchmark.

Cross-checking the range and keeping it current

Check the range two ways: compare the SDR:AE ratio with 1:2.4, with and without the top employer, and compare the range with an independent headcount or revenue-per-employee estimate. Then explain the gap in writing. Nobody has published how closely these methods usually agree, or how large a gap should trigger a recount.

The closest published guidance, from ORM, says to build both estimates and then reconcile the difference explicitly. I would add one discipline: decide your own tolerance before you look at the second number, and write it on the worksheet. Choosing the threshold after you see the gap defeats the purpose of the check.

The SaaStr staffing model is a useful reference for the headcount side. In its worked example, 16 fully ramped AEs correspond to 30 sales professionals. Even in a planning model, total sales headcount is nearly double the productive AE count. If your independent estimate starts from total headcount, expect it to sit well above an AE-based range.

Company A sits on the right of this matrix, with a ratio of roughly seven SDRs per AE. Its position on the vertical axis depends on the independent estimate. Until the title-variant re-run is done, the range goes out labeled as a floor-based estimate with one open item.

Keeping the estimate current is mostly a matter of discipline about queries. Decay of 2.1% a month adds up to 12% in six months, so an old count has drifted more than most haircut debates are worth. Because every count on the worksheet sits next to its exact query, a refresh means re-running those queries and comparing results. Saving the plain-English queries in Refolk makes that re-run a few minutes of work. When you refresh, look for three things: new AE start dates (a hiring burst lowers next year's capacity, especially in enterprise), new "now part of" employer strings, and AE profiles edited in the last few months, which point to reps who may leave soon.