HPNC — Health Insurance — Coconut Digital

The Situation

The most expensive period of the year. Every competitor active. Every target up.

This North Carolina health insurer sells insurance to under-65s in the state. Their busiest period is open enrollment — the roughly 10-week window between November and January when Americans can sign up for or change their health coverage. For any health insurer running Google Ads, this is also the most expensive period of the year. Every competitor is active, every budget is up, and CPCs rise accordingly.

The account had history. It had run through previous enrollment seasons with a CPA that hit as low as $34 during its best stretches, but with volatile spend patterns — campaigns paused and reactivated, budget deployed inconsistently — and structural issues that had never been fully resolved. Mixed keywords in ad groups. Irrelevant search terms burning budget. Ads without keyword relevance or clear CTAs. Demographic and device performance that had never been properly acted on.

Going into this enrollment season, the brief was straightforward: more enrollments, lower cost per enrollment, in a market where competition was only getting harder.

"Open enrollment is a useful test of whether an account is actually in good shape, because it's the one period when you can't hide behind a quiet market. Everyone's spending, CPCs go up. The only way to improve efficiency in that environment is to have done the foundational work before the season starts."

What the Audit Found

Seven issues. None complicated. Most just hadn't been done.

The value data was there. Nobody was acting on it. The highest-spending campaign had keywords like "blue cross blue shield nc", "affordable healthcare", and "medical insurance in North Carolina" grouped together — four different intents, one ad. None of the ads could be fully relevant to all of them.

High-CPA keywords absorbing budget unchallenged "Medicare supplement north carolina" had a CPA of $133. "Health insurance subsidy" had a CPA of $180. Both were running without bid controls that reflected what they were actually delivering.

High-value keywords being underserved "Health insurance nc" had a CPA of $19 but an impression share of only 35% — the account was showing on fewer than 4 in 10 searches for one of its best-performing terms. "Blue cross blue shield" had a $20 CPA and a 27% impression share.

Irrelevant search terms spending budget Terms like "from ready to complete to complete" had appeared 46 times. Words like "medicare" ($1,179 spent, $292 CPL), "advantages" ($2,259 spent, $488 CPL), and "companies" ($3,149 spent, $138 CPL) were pulling in searches that were never going to convert at a sensible cost.

Device performance ignored Mobile had a CPA of $96. Desktop had a CPA of $191. No bid adjustment had ever been applied. The account was paying the same for a desktop click that was three times less likely to convert.

Demographics not being used No audience bid adjustments in place. Ages 55–64 were converting at a CPA of $44. The bottom income tiers were converting at $43 — the most relevant audience for subsidised health plans — but weren't being prioritised.

Ads without keyword relevance or CTAs The highest-spending ad had a mismatched headline with no connection to the search term and no clear call to action. No filtering of low-intent clicks, no reason to click for the right ones.

What We Changed

Foundational work. Done before the season started.

STRUCTURE

Rebuilt campaign structure around intent Ad groups separated so that each one could serve a tightly relevant ad. Brand terms, competitor terms, and general health insurance terms treated as distinct audiences with distinct messaging.

BIDDING

Bids realigned with actual performance Manual bidding introduced. High impression-share targets set for the $19 and $20 CPA keywords that had been vastly starved of budget. Bids reduced on keywords where CPA data showed consistent underperformance relative to cost.

NEGATIVES

Negative keywords built from search term data Broad match negatives added at campaign level to block the categories of irrelevant traffic — out-of-state terms, job-search terms, international travel terms — that had been converting at multiples of the target CPA or not at all.

DEVICES

Device bids adjusted Desktop given a -33% bid adjustment to reflect the $191 vs $44 CPA gap. More budget effectively redirected toward mobile without reducing reach on the searches that mattered.

AUDIENCES

Demographic targeting applied Bid adjustments introduced across age ranges, genders, and income brackets based on historical CPA data. Lower-income segments — the most likely to qualify for subsidised plans — given higher priority.

ADS

Ads rewritten with keyword and CTA structure Headline 1 matched to the search term. Headline 2 a clear, commitment-signalling CTA to filter low-intent clicks. Description lines in sentence case, built around the specific audience and offer. New ads tested against old.

Three Months Later

85% more form fills and 26% lower cost per lead — measured against the same 3-month open enrollment window the year before.

Form fills up 85%. Clicks up 35%. Cost per form fill down 26%. Measured against the same open enrollment window the year before — same market conditions, same competition, same seasonal pressure. The account just handled it better.

The 35% increase in clicks combined with a 26% drop in cost per form fill means the account was both reaching more people and converting them more efficiently. The structural work done before the season — tighter ad groups, bid adjustments, negative keywords, demographic targeting — showed up in the numbers when the pressure was highest.

85% More form fills

26% Lower CPA

36% Spend increase

“

Open enrollment is a useful test of whether an account is actually in good shape, because it's the one period when you can't hide behind a quiet market. Everyone's spending, CPCs go up. The only way to improve efficiency in that environment is to have done the foundational work before the season starts — and to manage bids actively enough to respond when the market moves. The form fill volume and cost per lead numbers here are a direct result of fixing things that had been left unaddressed for a long time: structure, negative keywords, device bids, demographic targeting. None of it is complicated. Most of it just hadn't been done.

Melissa Ros · Founder

85% more leads. 26% lower CPA.

The most expensive period of the year. Every competitor active. Every target up.

Seven issues. None complicated. Most just hadn't been done.

Foundational work. Done before the season started.

85% more form fills and 26% lower cost per lead — measured against the same 3-month open enrollment window the year before.

Want an honest look at your account?