Tabellab

LinkedIn Audience Sizing: Job-Title Clusters, Company-Size and Industry Filters and an Audience That Is Neither Too Narrow nor Too Broad

Lesson 1 of 20Lab

All lessons
  1. LinkedIn Audience Sizing: Job-Title Clusters, Company-Size and Industry Filters and an Audience That Is Neither Too Narrow nor Too BroadLabFree
  2. What Attention Costs: CPM, CTR and CPC per Campaign, the Blended CPM and the CTR You Need for a Target Cost per ClickLab Locked
  3. From Click to Lead: Click-to-Lead Rate, Cost per Lead by Offer, a Target Check and the Blended CPLLab Locked
  4. Lead Gen Forms vs Landing Pages: CPL per Destination, the Form Saving, the Conversion Multiple and a Side-by-Side ChartLab Locked
  5. Lead Quality: Lead-to-MQL and MQL-to-SQL Rates, Cost per SQL and Why the CPL Ranking FlipsLab Locked
  6. From SQL to Pipeline: Cost per Opportunity, Pipeline per Dollar, Expected Revenue with Win Rates and SUMPRODUCTLab Locked
  7. Account Lists for ABM: Match Rate, Members per Matched Account, the Minimum Audience and How Many Accounts a List NeedsLab Locked
  8. Planning Forwards: From Budget and CPM to Impressions, Expected Leads, Expected CPL and Frequency in a Small AudienceLab Locked
  9. Who Got the Budget? Spend, Leads, CPL and Cost per SQL by Seniority with SUMIFS and a Conditional FormatLab Locked
  10. Lead Quality Check: Testing Leads Against the Target Profile with IF and AND, the On-Target Rate and the Cost per Usable LeadLab Locked
  11. Creative Fatigue in a Small Audience: Weekly CTR, the Trend with SLOPE, a Forecast and the Week to Refresh the AdLab Locked
  12. Testing Lead Gen Forms: Submit-Rate Difference, Pooled Standard Error, z-Score, p-Value and a Verdict per TestLab Locked
  13. Scaling the Budget: Average CPL vs the Marginal Cost of an Extra Lead with SLOPE, a Forecast and a Scatter ChartLab Locked
  14. The Sales Cycle: Days from Lead to Opportunity and to Close with DAYS, AVERAGEIFS by Destination and the MedianLab Locked
  15. Account-Based Reporting: Penetration per Account, Tier Goals with XLOOKUP, Multi-Threaded Accounts and Concentration with LARGELab Locked
  16. Allocating the Budget: Ranking Segments by Cost per SQL and Filling Each up to Its Maximum Spend with SUMIFS, MIN and MAXLab Locked
  17. What-If Grid: Expected Revenue per Dollar for Every CPL and Lead-to-SQL Rate, a Heat Map and the Highest Affordable CPLLab Locked
  18. Lead Scoring: A Weighted Score with SUMPRODUCT, Grades with IFS, the Cost per A Lead and Average Scores by DestinationLab Locked
  19. Budget Pacing: Running Totals, the Plan Line with DAY and EOMONTH, the Pace Index and the Month-End ProjectionLab Locked
  20. Final Project: LinkedIn Programme Dashboard by Campaign Group with CPL, Cost per SQL, Pipeline per Dollar, a Verdict and a Combo ChartLab Locked

You are not signed in — the lab works, but your progress is not saved. Sign up

Lesson 1 of 20 is free. Unlock all 20 hands-on labs, the final exam and your certificate.

Buy course — $9.90
Loading the spreadsheet…

Why audience size is the first number in a LinkedIn plan

Most advertising platforms let you start with a budget and discover the audience later. LinkedIn works the other way round, at least for business-to-business marketers. Its strength is that members describe their working lives in detail: their job title, their employer, how senior they are, which industry the company belongs to and how many people work there. A B2B advertiser can therefore say with unusual precision who should see an ad. The price of that precision is that the audience can become very small very quickly, and a small audience behaves differently from a large one.

That is why the first number in any LinkedIn plan is not the budget, the bid or the cost per lead. It is the size of the target audience: how many members match the description you have in mind. Everything else follows from it. How much budget the campaign can sensibly absorb, how often the same person will see the ad, how many leads a month are realistic and whether the plan needs one campaign or six all depend on how many people are on the other side.

This free lesson shows how to estimate that number in Excel before you open the campaign tool, using three inputs you can reason about: how many members hold the job titles you want, what share of them work at companies of the right size and what share of those are in your target industries. The example is Corvanta Systems, a fictional software company that sells dock and warehouse scheduling software to manufacturers and distributors with 200 to 5,000 employees. All companies, audiences, shares and results in this article and in the course are invented for learning, and every benchmark is illustrative rather than official platform data.

How LinkedIn targeting describes people

A consumer brand usually targets interests and behaviour. A B2B company sells to a role inside an organisation, so it targets professional attributes. The main ones are:

  • Job title: the exact wording members use, such as Warehouse Manager or Director of Operations. Precise, but titles vary a lot between companies, so one role needs a whole cluster of titles.
  • Job function: the broad department, such as Operations, Supply Chain, Information Technology or Purchasing. Wider than titles and more stable.
  • Seniority: the level in the hierarchy, from entry through manager and director to VP and the executive level.
  • Company size: bands by number of employees. A product built for mid-sized companies is wasted on a firm of ten or of fifty thousand.
  • Company industry: what the employer does, for example manufacturing, wholesale or logistics.
  • Company lists: instead of describing companies, you upload the names of the accounts you want. Lesson 7 of the course covers these account lists and their match rates.

Each attribute on its own describes a large group. The skill lies in combining them, because the combination decides whether you reach four hundred people or four hundred thousand. Attributes of different kinds narrow each other: a member has to match the title and the company size and the industry. Several values of the same kind widen the audience: Warehouse Manager or Distribution Center Manager.

Target on professional attributes only: role, seniority, company size, industry. Do not try to single out people by age, gender or other protected characteristics, directly or through proxies. It is unfair, it is restricted by platform policies and laws in many places, and for B2B it adds nothing: the job, not the person's private profile, decides who takes part in a purchase.

The core idea: every filter multiplies

When filters are stacked, each one keeps only a part of what the previous one left. That makes the arithmetic a chain of multiplications: target audience = members with the job titles × share at companies of the right size × share in the target industries. Take Corvanta's Logistics Manager cluster. About 380,000 members in the region hold one of the titles in the cluster. An estimated 27 % of them work at companies with 200 to 5,000 employees, and 44 % of those are in manufacturing, wholesale or logistics. The target audience is 380,000 × 0.27 × 0.44 = 45,144 members.

Every filter multiplies the audience down380,000 membershold a logistics manager job title in the region102,600 members× 27 %: at companies with 200–5,000 employees45,144 members× 44 %: in the target industries380,000 × 0.27 × 0.44 = 45,144 = target audience of the cluster
From 380,000 title holders to a target audience of 45,144: two filters remove 88 % of the starting group. Fictional data.

Two things about this chain are worth remembering. First, the shares are multiplied, never added. Adding 27 % and 44 % would suggest that 71 % of logistics managers are in the audience, when in fact it is 11.9 %. Second, the order of the filters does not matter for the result, because multiplication can be done in any order, but it matters for understanding: writing the steps down shows which filter costs you the most people, and therefore which one to loosen first if the audience turns out too small.

Multiplying shares assumes the filters are roughly independent: that logistics managers in manufacturing are about as likely to work at mid-sized firms as logistics managers elsewhere. That is rarely exactly true. Treat the product as a planning estimate, then compare it with the audience size the campaign tool forecasts once you enter the same filters.

Titles, functions and seniority: three ways to describe the same people

The same buyer can be described in more than one way, and the choice changes the audience size a great deal. Corvanta's buyer is the person who runs a warehouse or a distribution network. There are three common approaches.

  • Job titles only. The most precise description and usually the smallest audience. The risk is missing people whose title is worded differently: a Head of Logistics does the same job as a Logistics Director.
  • Function plus seniority. For example Operations at manager level and above. This catches unusual titles, but it also lets in roles you did not mean, such as an operations manager in a bank.
  • Titles grouped into clusters. A practical middle way, and the one used in this lesson. You collect the title variants of one role into a cluster, size each cluster separately and decide per cluster whether it earns its own ad set.

Working with clusters has a second advantage: the message can match the role. A warehouse manager cares about trucks queuing at the dock this afternoon, a VP of supply chain about cost per pallet across twelve sites, an IT manager about how the software connects to the existing warehouse system. One ad for all three would speak to none of them. But a separate ad set per cluster only works if each cluster is large enough to run on its own, which brings us back to size.

A fully worked example: Corvanta Systems

Corvanta's marketing team listed twelve job-title clusters across four functions. For each cluster the team estimated the number of members holding those titles in North America, the share at companies with 200 to 5,000 employees and the share in the three target industries. The table shows these inputs together with the two columns you calculate in the lab: the target audience and its status.

Job-title clusterFunctionMembers with these titlesCompany-size shareIndustry shareTarget audienceStatus
Warehouse ManagerOperations310,00028 %42 %36,456OK
Distribution Center ManagerOperations96,00034 %46 %15,014OK
Director of OperationsOperations420,00024 %31 %31,248OK
VP Operations / COOOperations265,00022 %30 %17,490OK
Logistics ManagerSupply Chain380,00027 %44 %45,144OK
Supply Chain DirectorSupply Chain150,00030 %41 %18,450OK
VP Supply ChainSupply Chain88,00029 %40 %10,208Too narrow
Transportation ManagerSupply Chain205,00026 %43 %22,919OK
IT ManagerIT1,450,00025 %21 %76,125Too broad
Director of IT / CIOIT340,00023 %20 %15,640OK
WMS / ERP Systems AnalystIT120,00031 %33 %12,276Too narrow
Procurement ManagerProcurement510,00026 %30 %39,780OK

The two calculated columns need one formula each, written for row 4 and copied down to row 15:

  • Target audience (F4): =ROUND(C4*D4*E4,0). Members × company-size share × industry share, rounded to whole members. ROUND(number, 0) rounds to zero decimal places, because 15,014.4 people do not exist.
  • Status (G4): =IFS(F4<$J$10,"Too narrow",F4>$J$11,"Too broad",TRUE,"OK"). The minimum sits in J10 and the maximum in J11. The dollar signs lock both cells so that every row compares with the same limits.

Walk through three rows to see the logic. The Warehouse Manager cluster starts with 310,000 members; 28 % at the right company size leaves 86,800, and 42 % of those in target industries leaves 36,456. That is comfortably inside the range. The IT Manager cluster starts far larger, at 1,450,000 members, because IT managers exist in every kind of company. Even though only 21 % of them work in the target industries, 76,125 remain. The VP Supply Chain cluster starts with only 88,000 members, and after both filters 10,208 are left.

Notice the Distribution Center Manager cluster. It has the highest shares in both filter columns (34 % and 46 %), because this title hardly exists outside the industries Corvanta sells to. Yet it ends with only 15,014 members, just above the minimum, because the starting group is small. A well-fitting title is not automatically a large audience.

Too narrow, OK or too broad

Is 10,208 members enough? Is 76,125 too many? There is no universal answer, but there are two forces every LinkedIn advertiser feels. Corvanta has turned them into a working range of 15,000 to 60,000 members per ad set. This range is the fictional company's own rule of thumb for its budgets. It is not a platform limit, and your own range will depend on how much you spend per ad set.

  • Too narrow. A small audience is used up fast. Within a few weeks the same members have seen the ad many times, the click-through rate falls and the cost of a thousand impressions rises, because few eligible members are available and other advertisers want them too. Lesson 11 measures this fatigue. Very small audiences may not deliver at all.
  • Too broad. A large audience delivers easily and often at a lower price per impression, but a growing part of the money reaches people who will never buy. The cost shows up later, as leads that sales rejects. Lessons 5, 9 and 10 measure it.
  • OK. Large enough to run for a quarter without wearing out, small enough that most members could plausibly be customers.

The IFS function fits this three-way decision. It takes pairs of a condition and a result and returns the result of the first condition that is true. The first pair asks whether the audience is below the minimum; if so, the answer is "Too narrow" and nothing else is checked. The second pair asks whether it is above the maximum. The final pair, TRUE, "OK", is the catch-all: if neither earlier condition applied, the audience is inside the range. The same result is possible with two nested IF functions, but IFS is easier to read.

Target audience by job-title clusterOKToo narrowToo broadWorking rangeWarehouse Manager36,456Distribution Center Manager15,014Director of Operations31,248VP Operations / COO17,490Logistics Manager45,144Supply Chain Director18,450VP Supply Chain10,208Transportation Manager22,919IT Manager76,125Director of IT / CIO15,640WMS / ERP Systems Analyst12,276Procurement Manager39,780Minimum 15,000Maximum 60,000
Nine clusters lie inside Corvanta's working range, two fall below 15,000 and one exceeds 60,000. Fictional data.

The chart makes the three groups obvious. Two clusters are too narrow: VP Supply Chain (10,208) and WMS / ERP Systems Analyst (12,276). One is too broad: IT Manager (76,125). The remaining nine are inside the range, although Distribution Center Manager (15,014) and Director of IT / CIO (15,640) are close to the lower edge and deserve a second look.

Totals by function with SUMIFS

Twelve clusters are too many to hold in your head, and too many to present to a manager who wants to know where the market is. The second half of the lab condenses them into the four job functions. The tool is SUMIFS(sum_range, criteria_range, criterion): it adds the cells of the first range for which the matching cell in the second range meets the criterion.

With the function names in I4:I7, the formula for the first function is =SUMIFS($F$4:$F$15,$B$4:$B$15,I4). Both ranges are locked with dollar signs so they stay in place when the formula is copied down; the criterion I4 is left relative so that it moves to I5, I6 and I7. The share of the total is then =J4/SUM($J$4:$J$7), again with a locked range for the total.

FunctionClustersTarget audienceShare of total
Operations4100,20829.4 %
Supply Chain496,72128.4 %
IT3104,04130.5 %
Procurement139,78011.7 %
Total12340,750100.0 %

The summary holds a small surprise. IT is the largest function with 104,041 members, 30.5 % of the total, although Corvanta thinks of itself as selling to operations people. The reason is the single IT Manager cluster, which contributes 76,125 on its own. Operations (100,208) and Supply Chain (96,721) are almost the same size, and Procurement, with one cluster, adds 39,780. All together the reachable market on the platform is about 340,750 members.

Step by step in Excel

If you want to rebuild the worked example in your own workbook, or follow along in the interactive sheet above, these are the steps:

  1. Enter the twelve clusters with their function, member count and the two shares in columns A to E, starting in row 4. Format the shares as percentages.
  2. Put the minimum (15,000) in J10 and the maximum (60,000) in J11. Keeping limits in cells instead of typing them into formulas lets you change them later in one place.
  3. In F4 type =ROUND(C4*D4*E4,0) and copy it down to F15.
  4. In G4 type =IFS(F4<$J$10,"Too narrow",F4>$J$11,"Too broad",TRUE,"OK") and copy it down to G15.
  5. List the four functions in I4:I7. In J4 type =SUMIFS($F$4:$F$15,$B$4:$B$15,I4) and copy it down to J7.
  6. In K4 type =J4/SUM($J$4:$J$7), copy it down to K7 and format the column as a percentage. The four shares must add up to 100 %.
  7. Check one row by hand with a calculator. If the hand result and the formula disagree, a reference points to the wrong column.

Common mistakes

The formulas are short, but the same errors appear again and again in real audience plans:

  • Adding shares instead of multiplying. Two filters of 27 % and 44 % do not keep 71 % of the audience. They keep 27 % of 44 %, which is under 12 %.
  • Forgetting a filter. Sizing by title and company size, then switching on an industry filter in the campaign tool, cuts the real audience to a fraction of the plan.
  • Typing limits into the formula. =IFS(F4<15000,…) works today, but when the range changes you must find and edit every copy. Reference the cells and lock them with dollar signs.
  • Missing dollar signs. Without $J$10 the reference slides down as the formula is copied, so row 5 already tests against J11 and the empty cell J12.
  • Testing in the wrong order. IFS stops at the first true condition. Put the catch-all TRUE first and every cluster is labelled OK.
  • Counting instead of summing. COUNTIF on the function column tells you that Operations has four clusters. It says nothing about how many people are in them. Use SUMIFS.
  • Double counting overlapping clusters. A member with the title Warehouse and Distribution Manager may match two clusters. The total of 340,750 is therefore an upper estimate; a combined audience in the campaign tool will be somewhat smaller.

How to read the result and what to do with it

A status column is only useful if each label leads to an action. For the three labels the usual moves are these.

  • Too narrow: merge the cluster with a neighbouring one. VP Supply Chain (10,208) and Supply Chain Director (18,450) together reach 28,658 senior supply chain leaders, who can share a message about network-wide cost. Alternatively widen a filter: the next company-size band up, or one more industry.
  • Too broad: add a filter or split the cluster. The 76,125 IT managers could be narrowed to those with warehouse or logistics systems skills, or to members at companies on an account list.
  • OK: give the cluster its own ad set and its own message, then let results decide. The later lessons of the course measure which of these clusters actually produce qualified leads and pipeline.
  • Near an edge: note it. A cluster at 15,014 is fine on paper but will show fatigue sooner than one at 45,000.

The function summary answers a different question: where should the budget go? If budget followed audience size, IT would receive the largest share. Corvanta will not do that, because IT rarely starts a warehouse project; it is brought in to approve one. Audience size tells you how many people can be reached, not how many should be. The value of each group is measured further down the funnel, in sales-qualified leads and pipeline, and the plan is adjusted as those numbers arrive.

Where the estimates come from

In this lesson the member counts and shares are given. In practice you gather them in three ways. The campaign tool forecasts an audience size for any combination of filters before a campaign is launched, so you can enter the titles alone, then add one filter at a time and note how far the number falls. Your own customer data shows which industries and company sizes actually buy. And sales colleagues know which titles sign the contract and which only attend the demo.

Forecast audience sizes are estimates and are shown in rounded ranges; they change as members update their profiles and as the platform changes its methods. Minimum audience sizes and other platform rules also change over time. The durable part is the arithmetic and the reasoning, which is what this course teaches. All numbers here are fictional and illustrative.

Keeping the plan in a spreadsheet instead of only in the campaign tool has a practical benefit. The sheet records why each audience was built the way it was, lets you compare alternatives side by side and gives you a baseline: when the campaign later reports who it actually reached, you can hold the report against the plan and see where reality differed.

What the rest of the course covers

Audience size is the starting point. The other nineteen labs follow the money from the first impression to the signed deal, always with Corvanta's fictional data and always in a live sheet that checks your formulas:

  • Costs (lessons 2 to 4): CPM, CTR and CPC, cost per lead by offer, and Lead Gen Forms compared with landing pages.
  • Quality and pipeline (lessons 5 and 6): from lead to MQL and SQL, cost per SQL, pipeline per dollar and expected revenue with SUMPRODUCT.
  • Account lists and planning (lessons 7 and 8): match rates, minimum audiences, and a forward plan from budget to expected leads and frequency.
  • Deeper analysis (lessons 9 to 15): breakdowns by seniority, on-target checks, creative fatigue with SLOPE, significance tests for form variants, the marginal cost of scaling, the sales cycle in days and account penetration.
  • Decisions (lessons 16 to 19): budget allocation with spend limits, a what-if grid, lead scoring and budget pacing with date functions.
  • Final project (lesson 20): a programme dashboard with a verdict for each campaign group and a combo chart.

Try it yourself

The interactive sheet at the top of this page contains the twelve clusters. Fill the target audience column, add the status, total the functions and calculate their shares; each cell is checked as soon as you enter it, and a hint is available if you get stuck. Then change the limits in J10 and J11 and watch the labels move. A company with half of Corvanta's budget might set the range at 8,000 to 30,000, and a different set of clusters would qualify.

Before you build any LinkedIn campaign, write the audience down as a multiplication: titles × company filter × industry filter. If you cannot estimate one of the factors, you have found the first thing to look up.

Frequently asked questions

How do I estimate a LinkedIn audience size in Excel?

Multiply the number of members who hold the target job titles by the share that works at companies of the right size and by the share in your target industries, for example =ROUND(C4*D4*E4,0). The shares are multiplied because each filter keeps only part of what the previous one left.

What is a good audience size for a LinkedIn B2B campaign?

There is no single correct number. It depends on budget, offer and how long the campaign should run. Many B2B advertisers aim for tens of thousands of members per ad set. The fictional company in this lesson uses 15,000 to 60,000 as its own working range, which is an illustrative rule of thumb and not a platform limit.

What happens if a LinkedIn audience is too small?

The same members see the ad repeatedly, the click-through rate falls, and the cost per thousand impressions tends to rise because few eligible members are available. Very small audiences may not deliver at all. Merging neighbouring title clusters or widening one filter usually solves it.

Should I target by job title or by job function and seniority?

Job titles are the most precise but miss unusual wordings. Function plus seniority catches more people but lets in roles you did not intend. A practical middle way is to group title variants into clusters, size each cluster and give the large enough ones their own ad set and message.

Why are the filter shares multiplied and not added?

A member must match every filter at the same time. If 27 % of title holders work at mid-sized companies and 44 % of those are in the target industries, the audience is 27 % of 44 %, about 12 % of the title holders. Adding the shares would wrongly suggest 71 %.

How do I label audiences as too narrow, OK or too broad in Excel?

Use IFS with the limits in two cells: =IFS(F4<$J$10,"Too narrow",F4>$J$11,"Too broad",TRUE,"OK"). IFS returns the result of the first true condition, and the final TRUE acts as the catch-all for audiences inside the range.

Keep going: LinkedIn Ads for B2B Lead Generation: Audience Sizing, CPM & CPL, Lead Gen Forms vs Landing Pages, Lead Quality, Pipeline & Account Lists

This free lab is lesson 1 of 20. Unlock the other 19 hands-on labs, the final exam and your certificate for a one-time $9.90.

  1. What Attention Costs: CPM, CTR and CPC per Campaign, the Blended CPM and the CTR You Need for a Target Cost per Click
  2. From Click to Lead: Click-to-Lead Rate, Cost per Lead by Offer, a Target Check and the Blended CPL
  3. Lead Gen Forms vs Landing Pages: CPL per Destination, the Form Saving, the Conversion Multiple and a Side-by-Side Chart
  4. Lead Quality: Lead-to-MQL and MQL-to-SQL Rates, Cost per SQL and Why the CPL Ranking Flips
  5. From SQL to Pipeline: Cost per Opportunity, Pipeline per Dollar, Expected Revenue with Win Rates and SUMPRODUCT
  6. Account Lists for ABM: Match Rate, Members per Matched Account, the Minimum Audience and How Many Accounts a List Needs
  7. Planning Forwards: From Budget and CPM to Impressions, Expected Leads, Expected CPL and Frequency in a Small Audience
  8. Who Got the Budget? Spend, Leads, CPL and Cost per SQL by Seniority with SUMIFS and a Conditional Format
  9. Lead Quality Check: Testing Leads Against the Target Profile with IF and AND, the On-Target Rate and the Cost per Usable Lead
  10. Creative Fatigue in a Small Audience: Weekly CTR, the Trend with SLOPE, a Forecast and the Week to Refresh the Ad
  11. Testing Lead Gen Forms: Submit-Rate Difference, Pooled Standard Error, z-Score, p-Value and a Verdict per Test
  12. Scaling the Budget: Average CPL vs the Marginal Cost of an Extra Lead with SLOPE, a Forecast and a Scatter Chart
  13. The Sales Cycle: Days from Lead to Opportunity and to Close with DAYS, AVERAGEIFS by Destination and the Median
  14. Account-Based Reporting: Penetration per Account, Tier Goals with XLOOKUP, Multi-Threaded Accounts and Concentration with LARGE
  15. Allocating the Budget: Ranking Segments by Cost per SQL and Filling Each up to Its Maximum Spend with SUMIFS, MIN and MAX
  16. What-If Grid: Expected Revenue per Dollar for Every CPL and Lead-to-SQL Rate, a Heat Map and the Highest Affordable CPL
  17. Lead Scoring: A Weighted Score with SUMPRODUCT, Grades with IFS, the Cost per A Lead and Average Scores by Destination
  18. Budget Pacing: Running Totals, the Plan Line with DAY and EOMONTH, the Pace Index and the Month-End Projection
  19. Final Project: LinkedIn Programme Dashboard by Campaign Group with CPL, Cost per SQL, Pipeline per Dollar, a Verdict and a Combo Chart

Buy course — $9.90 See the full course