Breaking Down Silos: What ACME’s Hub and Spoke Method Can Teach Us About Marketing
Blog by Kurt Hodges 9-4-2026
Most companies don’t have a data problem. They have a communication problem. In Cult of Analytics, Steve Jackson walks through a real example of this using a company called ACME. At ACME, hardware, software, sales, marketing, and customer support all worked as their own little islands, each with their own budget and their own goals (Jackson, 2016). Nobody was really talking to each other. Marketing didn’t know what sales was doing, sales didn’t know what was happening on the software side, and the company was losing money because of it.
ACME’s answer to this was something Jackson calls the hub and spoke method. Picture a team of analysts and data people sitting in the middle, that’s the hub. Around that hub are the different departments, or spokes, like marketing, sales, and support. The hub’s job isn’t to sit off by itself cranking out reports that nobody reads. Their job is to actually get in there with each department, figure out what’s really going wrong, and turn the data into something people can use.
What made this actually work wasn’t just some org chart shuffle. It came down to putting the right people in the right spot. Bill, the director of analytics, took his top analyst Jan and had her spend four out of five days a week sitting right inside the marketing department, along with a junior guy named Tommi who handled the tech side of things. Because they were actually there instead of just consulting from another building, they found problems nobody else had caught. Turns out marketing’s email system wasn’t even hooked up to the CRM. That one gap alone was costing the company around $10 million a year.
That’s really the big takeaway here for anyone running a digital marketing campaign. A campaign doesn’t happen in a bubble. Marketing might be watching clicks and engagement all day, but if none of that data gets over to sales, sales has no clue which leads are actually worth chasing. At ACME, once Jan and Tommi came up with a plan to automate email marketing based on white paper downloads, it wasn’t just a marketing thing. IT had to get involved to set up the tagging, the CRM team had to redo how they sorted prospects, and even the sales side got pulled in to build new white papers to feed the whole thing. One idea, four departments, and it only worked because everybody was talking from the jump.
I see the same thing play out in my own job managing suppliers in textile manufacturing. If purchasing cuts a deal without telling quality control, or production runs into a delay without letting sales know what was already promised to a customer, everybody ends up scrambling instead of just handling it ahead of time. It’s not really a data issue, it’s the same root problem, information staying stuck in one department instead of moving where it needs to go.
The numbers at ACME back this up too. That one email automation project brought in about $17 million in new sales over six months, and it happened because the analytics team quit treating each department like a separate customer and started treating them like partners. That’s the real value in sharing results across a company. Data doesn’t do much good sitting in a spreadsheet. It only matters once the right people see it, understand what it means for them, and actually have a reason to act on it together.
Reference
Jackson, S. (2016). Cult of analytics (2nd ed.). Taylor & Francis.
Why KPIs Only Work When They Tie Back to the Business
Blog by Kurt Hodges 9-17-2026
A lot of people throw around the term KPI like it just means “any number you’re tracking.” That is not really true. In Cult of Analytics, Steve Jackson makes a clear distinction between a count, a ratio, and an actual KPI, and the difference matters a lot more than it sounds (Jackson, 2016). A count is just a raw number, like how many visits your website got. A ratio is one number divided by another, like page views per visit. A KPI is different because it is “infused with business strategy.” In other words, it is not a real KPI unless it is actually tied to something the business is trying to accomplish.
Jackson breaks KPIs down into two types. There is the visionary KPI, which reflects the big picture goal of the whole company, and the tactical KPI, which is more specific to a department or team and feeds into that bigger goal. A good example from the book is ACME, the fictional company used throughout the text. ACME’s CEO decided the company needed to become number one or two in every market it competed in. That is a visionary KPI. It gave leadership a clear target and forced tough decisions, like selling off businesses that could not compete. But a goal like that means nothing to someone running a marketing campaign day to day. That is where tactical KPIs come in. ACME’s marketing team eventually built out specific tactical KPIs like conversion rate, bounce rate, and cost per user, all of which fed back into the same larger goal of driving more value at a lower cost.
For a number to actually count as a KPI, Jackson lays out four attributes it needs. It has to have a timescale, meaning someone checks it on a regular basis, whether that is daily, weekly, or monthly. It needs a benchmark, so people know what normal looks like and can spot when something is off. It needs an actor, meaning a specific person or team responsible for doing something with the information. And it needs an action attached to it, so if the number moves in the wrong direction, there is already a plan for what happens next. If any one of those four pieces is missing, what you have is just a number sitting on a report that nobody is really doing anything with.
The book gives a good real world example of this using MTV3, a media company in Finland. Their editors wanted to know when to change the content on their homepage. On its own, bounce rate is just a ratio. But once you attach a timescale (checked hourly), a benchmark (15 percent, with action triggered if it climbs past 18 percent), an actor (the page editors), and an action (swap out the content if a second metric confirms the drop in interest), it becomes an actual KPI that people can act on in real time.
This matters outside of digital marketing too. In my own work managing quality, maintenance, and engineering at a textile manufacturer, we run into the same problem all the time. A defect rate or downtime number means nothing on its own. It only becomes useful once someone owns it, checks it on a set schedule, knows what a normal range looks like, and has a plan for what to do if it moves outside that range. Without those four pieces, you are not managing anything, you are just watching a dashboard.
The real lesson from this chapter is that good KPIs are not really about the numbers themselves. They are about making sure every number tracked in a business ties back to a decision somebody is actually going to make.
Reference
Jackson, S. (2016). Cult of analytics (2nd ed.). Taylor & Francis.
Google’s New AI Summaries: A Good Starting Point, Not the Final Answer
Kurt Hodges 10/22/2026
Title: Google Ads gains AI dashboards built from text prompts
Author: Luis Rijo, PPC Land
Published: August 16, 2026
Link: https://ppc.land/google-ads-gains-ai-dashboards-built-from-text-prompts/
Citation: Rijo, L. (2026, August 16). Google Ads gains AI dashboards built from text prompts. PPC Land. https://ppc.land/google-ads-gains-ai-dashboards-built-from-text-prompts/
What the article is about
On August 10, 2026, Google announced new AI features for Google Ads and Google Analytics. Rijo (2026) breaks down four of them. Google Analytics now puts an AI summary at the top of the homepage that tells you what changed since your last login, like a traffic shift or a seasonal sales peak. Google Ads gets a prompt box and insight cards made for each business. Google Ads can also build dashboards from a typed request, and Analytics is next. Last, Google Analytics can compare your campaign results against anonymized averages from similar businesses.
The article doesn’t stop at the announcement. Rijo points out that Google gave no accuracy numbers, and that the announcement doesn’t mention confidence indicators or source citations. He also notes that putting the summary on the homepage makes it the default thing everyone sees, where before it was something you had to ask for. And none of the four features changes a live campaign. They summarize, show, and compare, and a person still decides what to do.
How this changes data analytics
I think the biggest change is speed. In our final project I spent a lot of time pulling numbers out of a spreadsheet, adding them up, and checking them. A tool that tells you what changed the moment you log in saves that time, and that means more people can get a quick read on their data, not just the ones who live in the reports.
It also changes the job. If the tool does the first pass, the human work moves to asking the right question and checking the answer. Jackson (2016) says to start with a business question, and a summary can’t pick that for you. The benchmarking feature connects to the KPI work from earlier in the course, where a number only means something when you compare it to a benchmark. But in my project the benchmarks came from the agency, and I couldn’t verify them. The same problem shows up here. The article says the announcement doesn’t explain how the similar businesses are picked, so you’d be comparing yourself to a group you can’t see.
The article also gives a good example of how people are using it. A paid media director quoted in it calls it a “directional check” and says the speed gives him “time to dig deeper” (Rijo, 2026). That’s how I’d want to use it too.
My position
I think this is a positive trend, with conditions. I like that it lowers the bar for people who aren’t analysts, and it gets the important changes in front of you quickly. I come from operations, where a good report shows you what needs attention first, and this does that.
My concern is the default part. When a summary is the first thing everyone reads, it’s easy to stop looking at the numbers underneath. In our project, the data file had figures that didn’t match each other. A nice summary built on those numbers would have looked just as confident as one built on clean data. The article also reports that 81% of the 1,448 U.S. consumers in one March 2026 survey were concerned about AI having access to their data (Rijo, 2026), so privacy is part of the picture too.
So my rule would be to use the summary to decide where to look, and then check the real numbers before anyone acts on it. Used that way, it’s a good tool. Used as the final answer, it’s a risk.
References
Jackson, S. (2016). Cult of analytics: Data analytics for marketing (2nd ed.). Taylor & Francis.
Rijo, L. (2026, August 16). Google Ads gains AI dashboards built from text prompts. PPC Land. https://ppc.land/google-ads-gains-ai-dashboards-built-from-text-prompts/