My working method
How I work with AI
I run my practice with Claude, Anthropic's AI model, working beside me on every project. It takes on the heavy data work: merging exports, reading crawls, scoring keywords and drafting first passes. I make the strategic calls and check every number before it reaches you. That's how one consultant delivers the depth you'd expect from a team, without the account managers or the handoffs.
client keyword libraries built on one shared method
blog posts planned across five editorial calendars
custom Claude skills I built and reuse across clients
routines that run on their own, daily or weekly
Where AI fits in my work
Here's where AI shows up in my work. Client examples are described by industry. Much of my current work runs through agency partners, and those relationships aren't mine to name.
SEO research and content strategy
For each client I build a keyword library that pulls rank tracking, Google Search Console and a full site crawl into one workbook, scored and grouped so a strategist can act on it. Claude does the merging and scoring. I set the rules and decide what matters. From there I build the editorial calendar and the action plan.
What it's turned up
- A sports advertising company's rank tracker showed 113 keywords in position 1. Only 3 were true first-place organic rankings. The other 110 were mentions inside Google's AI Overviews, which changes how you'd report success.
- For a healthcare products brand, the biggest opportunity sat in a product line the tracked keyword campaign had never targeted, on terms with low competition.
- Comparing Search Console against the crawl surfaced 85 forgotten legacy URLs that nothing on the site linked to. They were still earning roughly a quarter of the brand's organic clicks.
- For a precision manufacturer, I regrouped 413 tracked keywords by what the company can actually sell rather than by technique, so the report reads as a sales pipeline instead of a vocabulary list.
Analytics and tracking
I audit and repair GA4 and Google Tag Manager setups, then prove the fix on the live site instead of trusting the settings screen. Claude reads the full container export and walks the site with me, event by event, from product list to checkout.
What it's fixed
- On two online stores, a live walk-through confirmed that product lists, product views and clicks were all reporting correctly after the repair, down to the individual item.
- A "repeat purchasers" audience was filtering on order value instead of order count, so it treated about 90% of buyers as repeat customers. The rebuilt audience found 3.
- For a manufacturer's lead reporting, I separated real sales leads from contact-form traffic like job inquiries and vendor pitches, so the monthly lead count means what the client thinks it means.
Systems I reuse on every client
Anything I do twice becomes a system. My custom Claude skills carry the method from one client to the next: keyword library builds, content strategy workbooks, GA4 and Tag Manager setup, content refresh research, and a playbook for a long-running account. Every client gets the same structure and the same defensible method.
A few of the systems
- A content brief system split into four documents: a strategy note, a writer packet, an SEO pass and a CMS load sheet. Each person gets only what they need. Before anything's commissioned, a coverage check stops posts that would compete with a page the client already has.
- A monthly opportunity run that screens 15 or more content ideas against four tests. Is it on-service? Can it win against the sites holding the results today? Does it fit a real topic cluster with buying intent? Does the evidence hold up?
- A site evaluation framework organized into content, data, technical and authority tracks, with each criterion tagged for classic search, AI answer engines, or both.
- A quarterly strategy meeting template the room fills in together, so everyone leaves with the same decisions written down.
Automations that run before I sit down
Some of the most useful work happens on a schedule. These run on their own and report back, so my mornings start with a clear picture instead of a pile.
What runs
- Every morning, my inbox is sorted into what needs a reply and what doesn't, without anything being archived or deleted.
- A daily digest rebuilds my open-task list across every client, so nothing slips between projects.
- On Sunday evening, a week-ahead pack pulls my calendar, email and task list together into prep notes for the coming seven days.
- Scheduled research agents scan their sources each morning, confirm every result is still live before reporting it, and log what they found.
Building the business itself
I used Claude as a sounding board and production partner while setting up Ben Adams Consulting. It checked candidate business names against Florida's state business registry and domain availability, and helped me draft the business plan, contract templates, a break-even budget, outreach materials, case studies and the copy for this site.
How I keep AI honest
AI is fast and it sounds sure of itself, which is exactly why it needs checking. These are the rules I hold my work to.
A real example. In one build, the checks caught a traffic figure that came from the crawl tool's partial data instead of the full 12-month Search Console export. I corrected it in every file where it appeared.
What stays with me
Claude speeds up the research and the drafting. It doesn't set your strategy, talk to your customers or sign off on your work. I own the recommendations, and I review everything myself before it reaches you.
If your company has rules about AI and client data, tell me at the start and I'll work within them.
Want this kind of rigor on your account?
Tell me what you're working on and where it's stuck. I'll tell you plainly whether I'm the right fit.