Multichannel Marketing Without Directories Is Half-Done

I’ve been doing channel audits long enough to recognise the moment a client realises their stack has a hole in it. It usually happens about 40 minutes in, when I pull up their attribution report and ask why a third of their “direct” traffic searches for the brand name immediately after landing. That’s not direct traffic. That’s somebody who saw you somewhere else and typed your name into Google because they couldn’t remember the URL.

This article is the long version of that conversation, built around one client I worked with last year. The numbers are real (rounded for tidiness), and the principles travel further than I expected when I started.

The client that convinced me

A regional HVAC chain hitting a ceiling

The business: a five-location HVAC chain across the American Midwest, residential and light commercial. Roughly $14M in annual revenue, growing 8% year-on-year for three years running, then flat for two quarters. The owner, second-generation, cautious with money, sharp with numbers, called me because his marketing agency kept asking for budget increases and the phone wasn’t ringing any louder.

His exact words: “I’m spending more, and I’m getting the same. What am I missing?”

Their existing channel mix at $40K/month

Here’s what their spend looked like when I came in:

On paper, that’s a multichannel programme. In reality, it was four channels duct-taped together with two more that nobody owned. The blended CAC was hovering around $312, up from $244 eighteen months earlier.

Why their CAC kept climbing despite more spend

Three things were happening at once, and none of them were unusual.

First, Google Ads CPCs in the HVAC space had risen roughly 19% year-over-year, and their account history confirmed it. Second, Meta’s iOS attribution had been deteriorating since 2021, and the agency was reporting conversions that were, charitably, optimistic. Third, and this is the part that mattered, they had no presence on the platforms people actually used to compare HVAC contractors before calling one.

I asked the owner to walk me through how he picked his own dentist. He described checking three review sites and a local “best of” listicle before booking. Then I asked where his business showed up when somebody did the equivalent for HVAC. We Googled it together. He was on Google Business Profile. That was it.

Auditing the channel stack

What we found in their attribution gaps

I spent the first week running everything through an attribution audit. Tools used: GA4 (already installed but configured badly), CallRail for call tracking, Triple Whale for the paid social side, and a custom Looker Studio dashboard so the owner could see everything in one place without logging into six platforms.

The headline finding: 34% of “Direct / None” traffic in GA4 was preceded by a branded search within 30 days. That’s not direct. That’s assisted. We also found that 22% of inbound calls came from numbers we’d never tied to a campaign, meaning a fifth of their lead volume was floating around with no source attached.

The directory traffic they were dismissing as “noise”

Buried in the GA4 referral report were small trickles from Yelp, Angi, BBB, HomeAdvisor, and a regional contractor directory. The agency’s monthly report categorised these as “Other, minimal impact.” Combined, they were 6% of sessions. But those sessions had a 14% form-fill rate, compared to 2.1% from Google Ads.

Six percent of traffic was producing roughly 28% of qualified form-fills. The agency had been treating this as noise because it didn’t fit into the channels they were paid to manage. That’s measurement bias: if you can’t take credit for it, you call it noise.

Mapping discovery paths we hadn’t tracked

I ran a quick exit survey on the booking confirmation page for three weeks. One question: “Where did you first hear about us?” Free text. The results were instructive:

Word of mouth and vans I can’t bill for. But that 13% directory share was completely under-resourced. They had unclaimed or thin profiles on most of those platforms.

The fork in the road

Pause paid social or reallocate from search?

This is where most consultants make their money look smart and lose the client in the process. The instinct is to recommend something dramatic. Pause Meta. Cut display. Restructure Google Ads.

I didn’t, and here’s why: their search account was profitable at the unit level (just declining), and Meta was producing real demand at the top of funnel even if the platform’s reported numbers were inflated. Killing either would have created a hole faster than I could fill it.

Why we chose directories over a third retargeting layer

The agency’s pitch had been to add a third retargeting layer through a CDP and a programmatic vendor. Cost: roughly $7,000/month plus setup. Promised CAC reduction: vague.

I made a different argument. The exit survey showed that customers were already encountering the brand on directory sites, they just weren’t finding much when they got there. Building out the directory layer was cheaper, addressed an existing demand path, and didn’t require restructuring anything that was already working. The downside of being wrong was small. The downside of stacking another retargeting layer on top of weak attribution was a budget that nobody could justify in six months.

The 90-day test we proposed instead of a full pivot

I didn’t ask the owner to bet the farm. The proposal:

• Hold all existing spend flat for 90 days
• Add $3,500/month for directory presence (profile work, review generation, premium listings on two platforms)
• Defined success: blended CAC reduction of at least 15% by day 90, or we revert

That gave him a defined ceiling on risk. $10,500 over three months to test a thesis. He signed.

Building the directory layer

Picking 12 directories from a list of 60

I started with a master list of 60+ candidate directories pulled from BrightLocal’s citation tracker, Whitespark’s local citation finder, and my own working list. Then I filtered ruthlessly.

The criteria, in order of weight:

Final 12 (in priority order): Google Business Profile, Yelp, Angi, BBB, HomeAdvisor, Nextdoor (Business), Facebook (yes, it’s a directory now), Bing Places, Apple Business Connect, a regional Midwest contractor directory, an HVAC-specific industry directory, and a general business directory for the broader citation footprint. We used business directory for that last slot because it gave us a clean editorial listing without the spammy neighbourhood you find on a lot of free citation sites.

NAP consistency and the two weeks it took

NAP, Name, Address, Phone, consistency sounds boring because it is boring. It’s also the single most common reason directory work fails. Before we listed anywhere new, I audited their existing 47 citations across the web. We found:

• Three different phone number formats (parentheses, dashes, dots)
• Two slightly different business name spellings (one had “LLC”, one didn’t, one had “Heating & Cooling” vs “Heating and Cooling”)
• Four addresses with suite number variations
• Two zombie listings from a previous location they’d closed in 2019

It took two weeks of unglamorous work (Moz Local for the bulk update, then manual fixes for the platforms that don’t take feeds) to get to single-format consistency across all 47 existing citations. Only then did we start adding new ones.

Review velocity targets per platform

Reviews are what separate a directory listing from a tombstone. We set platform-specific targets:

The Yelp target looks small because Yelp will filter and hide reviews that look solicited; pushing harder there is counterproductive. Each platform has its own review etiquette and you ignore it at your peril.

What the numbers did

CAC dropped from $312 to $197

By day 90, blended CAC was $197, a 37% reduction against the $312 baseline. That beat our 15% success threshold by a wide margin.

The mechanism wasn’t mysterious. Search and Meta CACs barely moved. What changed was that the directory layer started producing leads at a CAC of roughly $74 (mostly because the cost was almost entirely the $3,500 setup investment plus management time, against a meaningful lead volume). Adding cheap leads to the blended pool dragged the average down.

The 23% lift in branded search we didn’t expect

Here’s what I genuinely didn’t forecast. Branded search volume, people Googling the company name directly, rose 23% over the same 90 days. We had not run a single brand awareness campaign.

What I think happened (and I’ll be honest, I can’t prove this with full certainty) is that the directory presence was acting as a memory aid. Somebody sees the company on Yelp while comparing options, doesn’t call immediately, then a week later remembers the name and Googles it directly. The directory wasn’t generating the demand; it was rescuing demand that would otherwise have leaked.

This matches what 73% have actually implemented one describes as the customer-centric reframing of multichannel attribution: stop asking which channel “owns” a conversion and start asking how channels work together across the journey.

Where directories underperformed our forecast

I’m not going to pretend everything worked. Three platforms missed targets.

HomeAdvisor produced leads, but the lead quality was terrible, heavily shared with competitors, lots of price-shoppers. We pulled the premium tier after 60 days and kept only the free profile.

BBB drove almost no traffic. Its value, as far as I can tell, is purely as a trust signal that customers check rather than discover through. We kept the listing because removing it would look strange, but I wouldn’t pay for it.

Nextdoor was a slow burn. The 15-recommendation target wasn’t hit until day 110. Worth doing, but don’t expect quick returns from a platform where the dynamic is neighbourhood-level word of mouth.

Principles that travel

Directories as connective tissue, not standalone channels

The most useful mental shift this engagement gave me: stop thinking about directories as a channel that competes with paid search or social. They aren’t a channel in that sense. They’re connective tissue, the surface where multiple channels’ worth of demand converges into a comparison decision.

Mailchimp’s product team has a framework I’ve borrowed: channels split into “drivers” and “destinations”. Email, ads, and social are drivers. Websites and checkouts are destinations. Directories sit in an awkward middle position. They receive driver traffic, work as a comparison destination, and then re-export traffic to your actual website. Skip them, and the comparison happens somewhere you have no presence (probably a competitor’s review page).

The “third touch” pattern across B2C services

Across roughly 30 B2C service businesses I’ve audited in the last four years, a consistent pattern shows up. The first touch is usually paid (search or social). The second touch is often organic search or a return visit. The third touch, the one that tips the decision, is disproportionately a directory or review site.

If you don’t show up at touch three, you don’t lose the customer to nothing. You lose them to whoever did show up.

When this approach actually fails

Directory layers don’t fix everything. I’ve seen them fail in three specific situations.

When the underlying service is bad. Directories magnify reputation. If your reviews trend below 4.0, building more directory presence accelerates the bleeding rather than slowing it. Fix operations first.

When the category isn’t searched comparatively. Some products (impulse e-commerce, certain B2B niches) don’t have meaningful directory ecosystems because nobody comparison-shops on directories. Pouring effort into a directory layer there is theatre.

When there’s no review velocity engine. A directory profile with 4 reviews is worse than no profile, because customers see the dead profile and assume the business is dormant. If you can’t sustain ongoing review generation, don’t build the listings.

Adjusting for different constraints

Running this play on a $5K monthly budget

The HVAC client had $40K/month to work with. Most businesses don’t. Here’s how I’d run a similar play at $5K total monthly spend:

At $5K, directories climb from 8.75% to 28% of total spend. That’s not because directories suddenly got better. It’s because at small budgets, the channels with high minimum-viable thresholds (programmatic, paid social, complex retargeting) stop working at all, while directories scale down gracefully.

The B2B SaaS version (G2, Capterra, niche lists)

The same principle works for B2B SaaS, but the directories are different and the economics are weirder. The B2B equivalent of Yelp is G2 or Capterra. The equivalent of a regional contractor directory is whatever niche list dominates your sub-category: Product Hunt for early-stage, Gartner Peer Insights for enterprise, Software Advice for SMB.

The big difference: G2 and Capterra are pay-to-play in a way that local directories generally aren’t. A premium Capterra listing in a competitive category can run $15K-50K/month. The math only works if your ACV justifies it (rule of thumb: don’t bother below $5K ACV, almost mandatory above $25K).

The review velocity problem is also harder. B2B buyers leave fewer reviews than consumers, and the review platforms know it, so they offer gift cards in exchange. This is allowed, disclosed, and weirder than it sounds.

Compressing the timeline to 30 days

Sometimes you don’t have 90 days. A new GM joins, a quarterly board meeting looms, a seasonal window closes. If I had to compress this play to 30 days, here’s what would change and what wouldn’t.

What stays: the NAP audit and consistency work. Skip this and everything else underperforms. Two weeks minimum, even compressed.

What changes: instead of 12 directories, I’d pick 4: Google Business Profile, Yelp, the single biggest vertical-specific directory, and one general directory. Concentration over coverage. Review velocity targets get aggressive (use a tool like Birdeye or Podium to automate the post-service ask within 24 hours). Skip the platforms with slow burn (Nextdoor, BBB, niche industry lists) entirely; they won’t show meaningful results in 30 days regardless of effort.

Realistic 30-day result: 10-15% CAC improvement rather than 37%. Not as dramatic, but defensible.

The harder constraint: organisational, not budgetary

The constraint I find hardest to design around isn’t budget or timeline. It’s organisational. If your paid media is run by an agency, your SEO by a contractor, your reviews by the operations team, and your CRM by a fractional consultant, nobody owns the directory layer. It falls into the gaps between people who all assume someone else is handling it.

The HVAC client worked because the owner made one person (his marketing coordinator) the single point of accountability for everything that touched directories: listings, reviews, NAP consistency, response to negative reviews. Not the agency. Not me. An internal person whose job depended on it.

If you can’t create that ownership, the play doesn’t work. I’ve watched well-funded directory programmes wither because three vendors each thought review responses were the others’ job. The platforms decay, the reviews go unanswered, the listings drift out of NAP sync, and within nine months you’re back where you started, except with worse public reviews.

Where I’d start tomorrow if I were you

Pull up your GA4. Filter to the last 90 days. Look at the referral report and find every directory-style source sending you traffic. Now check the conversion rate of those sessions against your paid search conversion rate. If the directory traffic converts better, and in 8 out of 10 audits I run, it does, you have your answer about where the next $1,000 of monthly spend should go.

Then check whether you’ve actually claimed and filled out those profiles, or whether your business has been earning conversions on autopilot from listings you’ve never touched. That gap, between accidental presence and intentional presence, is usually where the first 20% of CAC reduction lives. The remaining 17% takes the work I described above.

Get the connective tissue right, and the channels you’re already paying for start performing better. Skip it, and you’ll keep wondering why more spend isn’t producing more leads. The math doesn’t change just because the answer is unsexy.