GuavaBox Story - Dramatically Scaling Traffic Using Batching

This is an Agency Breakthrough episode (ZenPilot’s other show) with no outside guest. Kuba Grajcar hosted and interviewed Gray MacKenzie about GuavaBox, the inbound marketing and web design agency Gray launched with Andrew Dymski out of a college dorm room at the start of 2011. It’s built around one breakthrough: weekly marketing batch days, and what they did to the agency’s traffic and pipeline.

A year as digital yes men, then a bet on HubSpot

Gray described GuavaBox’s first year as a year in the wilderness as “digital yes men”: if you had money, they’d say yes, figure it out, and chronically undercharge. Four partners started it; nine months in there were three. Kuba asked what problem they were solving.

The first problem was we made $31,000 as a company - not made in terms of profit, made in terms of revenue - in 2011. I was the only one who was full-time, fortunately, but still my paycheck, or my distributions that year, were $16,000 pre-tax.

(Listen at 05:42)

There was no site traffic, no sales process and no funnel. In mid-2012 they stopped being generalists.

We made the craziest investment. We spent more than 25% of our prior year’s entire revenue of the company to invest in HubSpot and say, we’re going to go all in in terms of inbound marketing. These are the services that we’re going to start packaging and offering to our clients.

(Listen at 07:31)

They’d use it on themselves first. Gray said he and Andrew have always been on the responsible, less aggressive end of the entrepreneur scale, and they wanted inbound working for GuavaBox before charging a client three grand a month for it. The pattern they were trying to escape is the stair step most agencies know: sell, go heads down delivering for two months, realize there’s no next project, go back to selling. “That’s exactly the hamster wheel that we were stuck on.”

What a Wednesday batch day looked like

So every Wednesday, for the entire morning, from when we got in in the morning until 1pm, everyone on the team would just spend the entire time writing, and primarily blogging. We’d have a prioritized list coming into the day, we’d have done the SEO research to figure out what we were going after, and you just write as many posts as you could get through. Some days that was half of one and some days that was three blog posts that you could write.

(Listen at 11:49)

SEO was the channel, and it paid off within the nine months they kept the schedule.

We went from a couple hundred site visitors a month… to over 15,000 visitors a month in traffic, which blew up our pipeline. Our site was converting at about 3% from visits to contacts.

(Listen at 13:03)

Pain keywords, solution keywords, and a Pittsburgh post

Kuba asked how they picked topics. Gray’s grid had two buckets: pain-based keywords (the open-ended “how do I solve this” searches) and solution-based keywords, where the searcher already knows what they want. Nobody was searching for inbound marketing yet, but plenty of people wanted SEO services, and GuavaBox went local with it.

We wrote these posts. One was called Five Lies Pittsburgh SEO Companies Like to Tell… Well, that did great, because one, it was the number one result for Pittsburgh SEO company very quickly. And then two, the click-through rate on that was extremely high. Who’s written the article on five lies that these SEO companies like to tell?

(Listen at 17:14)

They also mapped content to funnel stages, including a top-of-funnel post on Facebook contest apps that still ranked near the top when Gray checked during the recording.

Plan it, leave time to publish, celebrate

Kuba asked for a recipe. Gray gave three parts.

Number one, somebody needs to plan it. So we would rotate this role between three of us who would take point, but basically, what’s the priority order?… That does not need to be a whole group effort to go do that. Someone can easily do it, and if it’s slightly wrong, that’s fine.

(Listen at 23:09)

Two: keep the last 45 minutes to an hour for editing and hitting publish, with a writer, a proofreader and a publisher assigned per post. Three: end with a quick round of updates on what shipped and what’s in the hopper.

You’re celebrating the fact that we showed up, we did our work for this week. We can now check out, and if we don’t touch marketing again until next Wednesday, that’s totally fine. We did the minimum standard of, we can call it a win because we all showed up and we did this.

(Listen at 25:34)

Every post ran through the same five-step checklist: writer, editor, optimizer, publish, promote. One lesson learned the expensive way: a Getty Images letter over an unlicensed featured image that cost $1,200.

Why it slowed down, and what Gray would do today

Kuba asked what made them stop after nine months. They didn’t stop; the block shrank from four or five hours to two.

The biggest reason for that was we’re kind of swimming in more leads than we knew how to handle at that point.

(Listen at 27:24)

Delivery became the next thing to break, which led to systematizing it and eventually building DoInbound. Starting over today, Gray said he’d dedicate a full day or two half days a week and scale back once results came. On tools, his advice was not to let the tool be the hang-up: tasks live in ClickUp, writing happens in Google Docs, and the real payoff is the morale of being in it together.