Why are my LinkedIn ads not converting? Route it by the broken number

Find the broken number before you touch a single setting

A campaign that is not converting has exactly one gate failing first, and that gate is the only thing you are allowed to work on this week. Everything else in the account is either fine or unreadable until the first failure is cleared. This is the part the standard six reasons article cannot give you, because a list has no order and a failure does.

The order is fixed for a mechanical reason. Each number in a paid funnel is calculated from the one above it, so a broken number upstream makes every number downstream meaningless rather than merely worse. A click-through rate calculated on 40,000 impressions is a fact. The same rate calculated on 4,000 impressions is a rumour.

Gate one is delivery, and it disqualifies most diagnoses before they start

Two checks, both of which take a minute. First, is the campaign spending close to its daily cap on most days? Second, how many impressions did each person in the audience receive across the flight? Divide total impressions by the forecast audience size to get that second number.

Consistent underspending against the cap means delivery is being constrained rather than chosen. The usual causes are a bid set below what the auction is clearing at, an audience too small to fill the budget, or an ad the system is reluctant to serve. Raise the bid in small steps before you do anything else, because a constrained campaign produces distorted numbers at every gate below.

The impressions per person number is the one nobody calculates. Under three across the flight, the campaign never reached anybody often enough to be recalled, and the click-through rate you are staring at is measuring first sightings by strangers. The fix is to cut the audience, not the creative, and the argument that this reduces opportunity is answered by pointing at the number.

Gate two is the click, and 0.4% is a triage line rather than a benchmark

With delivery healthy, a click-through rate under 0.4% means the audience you paid to select is declining your message. Treat 0.4% as a routing device, not a verdict. The published click-through benchmarks for LinkedIn disagree with one another by more than the difference you are trying to detect, because they average across formats, countries, seniorities and years that have nothing to do with each other. Replace the line with your own account's trailing median as soon as you have twelve weeks of history.

The useful question at this gate is not how to write a better ad. It is whether the failure is match or message, because the two have opposite fixes and they look identical on a summary screen.

The discriminator at gate two. Both tests need enough delivery on each variant to be readable, which is why gate one runs first.

Most teams skip the discriminator and rewrite the creative, because rewriting is satisfying and re-targeting an audience feels like admitting the strategy was wrong. The rewrite has roughly a fifty percent chance of addressing the actual cause, which is a coin toss dressed as a decision.

Gate three is the conversion, and low completion is an offer problem

Clicks arriving and leads not arriving means the destination is asking for more than the ad earned. Use 8% click to lead as the triage line for a landing page and treat it the same way as the click gate, as a router rather than a scoreboard. Native lead forms normally run far higher because the fields are prefilled and the friction is close to zero, so never compare the two against a single number.

Split the gate into two measurements before you decide anything. Look at how many sessions start the form, and then how many of those finish it. The two failures point in different directions and the combined conversion rate hides both.

Reading gate three properly requires two numbers rather than one.

The single most common finding at this gate is a mismatch between what the ad promised and what the page offers. The ad sells a specific idea, the page sells the company. Lead Gen Forms and landing pages fail differently, and choosing between them is a decision about which failure you would rather diagnose.

Gate four is qualification, where healthy form fills meet an empty calendar

Leads arriving and meetings not arriving is a qualification failure with three possible causes, and you can separate them in twenty minutes without changing a single setting. Pull the last twenty five leads into a spreadsheet and tag each one on two columns: does the job title match your target, and what did they say or not say in reply.

Reading the tags from your last twenty five leads. The pattern in the two columns names the cause.

Notice that only one of those four rows is fixed inside Campaign Manager. This is why gate four failures survive so long. The dashboard is where people look and the cause is somewhere else, so month after month the creative gets rewritten and the calendar stays empty.

If the tags come back scattered and junior, junk leads have their own cure and it starts with the offer rather than with more filters.

Run the whole route in ten minutes

Pull five numbers in this order and stop at the first one that fails. Do not open the creative until you have all five, because looking at the ads first is what makes people diagnose from taste instead of arithmetic.

Ten minutes of this beats a week of creative review, and it also gives you something to say in the meeting that is not an apology. Naming the failing gate turns a vague problem into a specific one with an owner.

A worked example, illustrated rather than real

The following uses invented numbers to show how the route runs. It is an illustration, not a client and not a case study. No result is being claimed for anybody.

A software company spends $4,000 in a month at an assumed $35 CPM, which buys about 114,300 impressions. The audience is 50,000 people. Click-through rate reads 0.31%, giving roughly 354 clicks. The landing page converts 13%, giving 46 leads at about $87 each. Two meetings were booked. The team's instinct is that 0.31% is the problem and the creative needs rewriting.

Run the route instead. Gate zero passes, because the CRM shows 46 leads too. Gate one fails: 114,300 impressions divided by 50,000 people is 2.3 impressions per person, which is under the floor of three. The route stops there. The 0.31% click-through rate is not a message verdict, it is a measurement taken under conditions where most of the audience saw the ad twice at most.

The prescribed change is to cut the audience from 50,000 to roughly 20,000 and keep everything else identical, which lifts impressions per person to about 5.7 on the same budget. That is not a promise that click-through rate will improve. It is the change that makes the next reading of click-through rate mean something, which is the only honest thing anybody can offer at gate one.

The purpose of the route is not to be right about the cause on day one. It is to stop you from spending a quarter answering a question the data was never able to answer.The house position

Three fixes people try first that almost never work

These three appear in nearly every emergency review, and all three fail for the same underlying reason: they change something downstream of the gate that is actually broken.

Rewriting the creative while delivery is failing

New creative resets whatever delivery history the campaign had accumulated and produces a fresh unreadable sample against the same overly wide audience. The team then compares two unreadable samples and picks a winner from noise, which is worse than not testing at all because it produces a confident wrong conclusion.

Adding targeting filters to improve lead quality

Stacking filters shrinks the audience below what the budget needs and raises CPM at the same time, since the remaining inventory is scarcer and more contested. Quality problems that come from the offer are not solved by filters, and the campaign is now both narrower and more expensive.

Switching bid strategy to fix a conversion problem

Bidding controls what you pay for attention. It has no influence on what happens after the click. Changing it when gate three or gate four is failing changes your cost per click and leaves the actual failure exactly where it was, with a new variable now confusing the comparison.

When the ads are fine and the problem is downstream

Sometimes every gate clears and the pipeline still does not move, which means the ad account was never the constraint. Three downstream causes account for most of it, and none of them are visible in Campaign Manager.

• Response time. A lead that waits two days has cooled, and no bidding change recovers it. Assign an owner and an agreed response window before you raise the budget.
• The definition of qualified. If sales and marketing never wrote it down, every review becomes an argument about definitions and the ads lose that argument regardless of what they produced.
• What the buyer finds when they check you out. Attention bought by an ad gets verified against your people and your page, and if there is nothing to find the click was rented rather than earned.

That last one is slower to fix and it is the reason paid and organic are not separate budgets in practice. A written employee advocacy policy is the unglamorous mechanism behind having something for a curious buyer to find, and it costs no media money at all.

None of this excuses a badly run account. It does mean that a campaign passing all five gates and still producing nothing is telling you something true about the business rather than something false about the channel.