What We Learned from the First 100 Businesses on KOIRA

We didn't plan to write this post. But after onboarding our first 100 businesses, the patterns became too consistent to ignore — and too useful to keep internal.

These are not feel-good stories about growth. This is a post-mortem on what actually happens when small and medium business owners try to take control of their own marketing. Some of it is encouraging. Some of it is uncomfortable. All of it is real.

The First Thing Every Business Got Wrong

Before any automation was set up, before a single piece of content was drafted, we asked every business owner a simple question: Describe your brand voice in three sentences.

Fewer than 20% could do it.

Not because these owners didn't have a voice — they absolutely did. You could hear it the moment they started talking about their business. The problem was that it had never been externalised. It lived entirely in their heads, and sometimes in their best employees' heads, but nowhere that a system — or a new hire, or a freelance writer — could access it.

This is the foundational problem of SMB marketing. It's not a budget problem. It's not a tools problem. It's a documentation problem. The business exists. The personality exists. But the moment you try to scale any kind of content production, you hit a wall because nobody wrote any of it down.

The businesses that moved fastest through onboarding were the ones who had some version of a brand voice document — even a rough one. A paragraph of "we always say X, never say Y." A saved email they were proud of. A Google Doc with their elevator pitch. Anything.

If you don't have that, start there. Not with tools. With words.

The Consistency Finding Nobody Expected

We expected the businesses with the best content to get the best results. We were wrong.

The clearest predictor of early SEO and engagement gains was publishing frequency, not quality. Businesses that published two pieces of useful, moderately-polished content per week consistently outperformed businesses that published one exceptional piece per month.

This runs counter to everything you read in content marketing guides. "Quality over quantity" is repeated so often it has become unexamined dogma. But for SMBs competing in local and niche markets — where the bar is genuinely low — showing up consistently is a meaningful competitive edge.

The algorithm rewards freshness. Customers reward presence. A business that answers ten common questions in ten straightforward blog posts will outrank a competitor with one beautifully crafted pillar page, at least in the short and medium term.

This doesn't mean quality is irrelevant. It means that for most small businesses, the bottleneck is not quality — it's volume. You can't polish what you haven't published.

The Approval Queue Was Not What We Expected

We built the approval queue as a trust mechanism. The idea was that business owners who weren't ready to fully automate content could review and approve everything before it went live. We thought it would be a transitional feature — something people used early and then turned off as their confidence grew.

That's not what happened.

Owners loved the approval queue. Not as a temporary crutch, but as a permanent fixture. After months of using the platform, most owners still wanted to see everything before it published. Not because they distrusted the output — the quality complaints were actually very low — but because they wanted to feel connected to their marketing.

This was a genuine insight. These owners hadn't hired a marketing team precisely because they didn't want to hand their brand to someone else. Giving them automation with a final say was fundamentally different from giving them automation that ran without them. The approval step, even when they approved 95% of content unchanged, gave them a sense of ownership that kept them engaged.

The lesson for anyone building marketing systems — automated or otherwise — is that control and convenience are not opposites. The best tools give you both.

The NAP Problem Was Everywhere

Nearly 60% of the businesses we onboarded had inconsistent NAP data — name, address, and phone number — scattered across the web. Wrong suite numbers. Old phone numbers from three moves ago. Business names with and without "LLC." Addresses formatted differently on Google, Yelp, and their own website.

This matters because local SEO relies heavily on citation consistency. When Google sees five different versions of your address across twenty directories, it loses confidence in your listing. That lost confidence shows up as lower local pack rankings.

Most of these owners had no idea. They'd moved, changed their number, updated their website — but never thought about the downstream directories that had cached their old information.

Fixing NAP consistency was, in many cases, the single highest-ROI action we took in the first thirty days. No new content. No campaigns. Just making sure the business existed accurately on the internet. Local search performance improved for nearly every business where this was corrected.

If you haven't audited your citations recently, Moz Local and BrightLocal are both solid tools for finding the discrepancies.

The Automation Overreach Problem

About a third of businesses came in wanting to automate everything immediately. Email sequences, social posting, blog content, review requests, Google Business Profile updates — all of it, from day one.

Almost all of them stalled.

The problem wasn't technical. The problem was that automation without a foundation produces noise. If your brand voice isn't documented, automated content sounds like it was written by a committee that never met your business. If your contact list isn't segmented, automated emails annoy people instead of helping them. If your Google Business Profile isn't complete and accurate, automating posts to it just draws attention to a broken listing.

The businesses that got the best results from automation were the ones that automated one channel first, got it working well, and then expanded. Email before social. Social before ads. Core content before distribution.

This is counterintuitive when you're looking at a platform full of capabilities. The temptation is to turn everything on. The discipline is to turn on one thing and make it actually work.

What Local Signals Did That Nothing Else Did

For the 70% of our first 100 businesses that served a local market — trades, retail, food service, professional services, health and wellness — local signals delivered the fastest measurable ROI of anything we tracked.

Specifically: Google Business Profile posts, active review response, and locally-targeted landing pages. Not paid search. Not social media. Not email. Local organic signals.

This is partly because local SEO remains underleveraged. Most small businesses in local markets are not doing it systematically. Claiming a Google Business Profile is not the same as actively using it. Posting weekly updates, responding to every review within 24 hours, adding photos regularly, keeping hours accurate — these actions compound. And because most competitors aren't doing them, the baseline you're competing against is low.

For service businesses in particular, we saw ranking improvements in local pack results within 60 to 90 days of consistent GBP activity. No technical SEO changes. No link building. Just showing up on the platform Google gives you for free.

The Voice Ingestion Moment

One of the more technically interesting findings came from businesses that spent time on voice and context ingestion — feeding the system with examples of their own writing, their tone, their specific vocabulary and phrases.

The content output quality difference was significant. Not marginal — significant. Businesses that put in two hours of voice documentation produced AI-assisted content that their customers recognised as authentically theirs. Businesses that skipped it produced content that was technically correct but generically forgettable.

This gap narrowed over time as the system learned from approved edits and feedback. But the starting point mattered. The businesses that invested in voice documentation early didn't have to spend the next three months correcting outputs that didn't sound like them.

The takeaway is simple: garbage in, garbage out applies to brand voice just as much as it applies to data. If you want an AI system to sound like you, you have to tell it what you sound like.

The Honest Number

Of the first 100 businesses, roughly 65% were seeing measurable positive results — in traffic, in leads, in revenue, or in time saved — within 90 days.

That means 35% weren't. And we've looked hard at why.

The common thread in the businesses that didn't see results in the first 90 days was not the tools. It was engagement. Owners who checked in weekly, approved content promptly, gave feedback on what was and wasn't working — they saw results. Owners who set things up and then disengaged, assuming it would run without their input, found that it mostly didn't.

Marketing automation is not a set-it-and-forget-it system. Not at the SMB level. It's more like hiring a very capable assistant who still needs to know what you want, get your sign-off on important things, and hear from you when they've got it wrong. The owners who treated it that way got the most out of it.

That's not a product limitation. That's just how good marketing works.

What We're Changing Because of This

Everything above has changed how we think about onboarding and tooling. A few specifics:

• Voice documentation now comes first in onboarding, before any automation is configured.
• NAP audit is now a standard day-one action, not an optional add-on.
• The approval queue is now a first-class feature, not a transitional one.
• We've built better guidance around single-channel-first automation to prevent the overreach problem.

The first 100 businesses taught us more than we expected. The next 100 will teach us more still.