Kevin Dees shares practical advice on business technology, marketing and AI-assisted custom software. The posts discuss custom development and tool choices for Shopify clients, alongside guidance on LinkedIn and business attention.
Kevin Dees is most useful for practical guidance on AI-assisted custom software, Shopify tool choices, and business technology tradeoffs.
Best for
A Shopify app vendor or development agency explaining when custom, AI-assisted development beats off-the-shelf tools.
Not for
A consumer brand wanting high-reach awareness, paid sponsorship package, or segmented demographic audience.
Key takeaways
AI can structure company research, but the output still needs checking.
Choose software by fit, not feature count, weighing off-the-shelf apps against custom development.
Attention is a business issue, not just a marketing one.
Check first: Confirm rate card, media kit, sponsored examples, inquiry route, and post-level analytics.
Main ideas
AI can structure company research, but the output still needs checking · Choose software by fit, not by feature count · Attention belongs to business operations, not just marketing
Kevin Dees shares practical perspectives on business technology, custom software, AI, and marketing, mainly through LinkedIn posts.
Their examples range from using AI to organize research about a company to deciding when a Shopify workflow may justify a small custom app.
A podcast recap and a post about business attention show a broader interest in how people use professional platforms and make day-to-day decisions.
Dees connects technology choices to the business problems they are meant to solve.
In the Shopify post, they weigh the excess features of some off-the-shelf apps against the cost of custom development, rather than presenting either option as universally better.
The AI-company-audit post offers prompts readers can try and uses AI for objective brand audits.
Across these posts, Dees favors practical questions and tradeoffs over technology for its own sake.
Readers interested in business software decisions, AI-assisted research, or more purposeful professional-platform activity may find the work useful.
Dees writes about building and choosing software for businesses, mostly around Shopify and custom development, and about how AI changes both how companies research each other and what they should build.
Read more
Two priorities run through the posts: judge a tool by whether it fits the actual workflow rather than by how many features it has, and treat attention as an operating decision rather than a marketing metric. They also publish reusable prompts for using AI in company research; verify a model's assessment against source material.
AI can structure company research, but the output still needs checking
Dees asked Google Gemini to research the firm, Robojuice, and report whether they should hire it, then published the three prompts used: one to audit credibility and contradictions, one to check market positioning against the industry vertical, and one to extract exactly three pros and three cons.
Read more
The prompts are reusable; verify the generated assessment against source material before relying on it.
The Gemini output described Robojuice as a boutique consultancy with direct founder access, cited TypeRocket as an open-source framework behind more than 10,000 professional platforms, and listed limited scale, key-person dependency, and premium pricing as cons.
Dees describes the 90/10 app problem: a client evaluates apps that cover 90% of features they do not need in order to get the 10% they do.
Read more
They argue AI has made small custom Shopify apps more practical for narrow workflows because the platform is well documented and current models handle its APIs well. The tradeoff is real development cost against less feature clutter and technical debt, with software shaped around how the business actually works.
Rewind and Webgility are named as strong apps for complex needs such as backups and accounting integrations, while a client's very specific workflow is where a small custom app can beat an off-the-shelf one.
Attention belongs to business operations, not just marketing
Dees argues that attention is part of how a business runs, not a marketing tactic.
Read more
It goes to threats such as a spoofed e-commerce site, to client leads that change a team's workload, and to the relationships that need you.
Promotion, socializing, and ads can add noise rather than cut through it.
Deciding when to pivot means weighing what currently has your attention against what deserves it.
A spoofed storefront and a workload-changing client lead are concrete operational cases, not promotional ones, and both compete for the same attention as a campaign.
Relaying points from a Website Leader Podcast conversation with guest Richard Bliss, Dees passes on three arguments: LinkedIn prioritizes user activity over ads because subscriptions drive its revenue; post from a personal account rather than a business page; and the algorithm amplifies conversation, so meaningful comments matter more than broadcasting.
Read more
These are Bliss's points as Dees presents them, framed for leaders under pressure to keep up with a fast-moving digital landscape.
The specific routine is 30 minutes a day leaving meaningful comments on other people's posts, with conversation treated as what the platform rewards.
Dees flags updates to tools relevant to the clients they advise, showing how recommendations grow out of hands-on client work rather than general endorsement.
Read more
Ask whether a paid relationship exists; none is documented here.
After talking with West Jones at Atlas Local, Dees praised Supermoon's Shopify integration, one-click Gmail integration, and AI auto-tagging controls, and said they would check it out for Shopify clients they handle custom development for.
Open this for prompts to ask AI to organize public-company research.
Read more
Dees shares three prompts for examining a company’s credibility, market positioning, and strengths and weaknesses, then shows the assessment Gemini generated about Robojuice. The example makes the prompts easy to try and shows why an AI-generated assessment is a starting point for research, not independent proof of its claims.
For brands planning a campaign
Could inspire a hands-on campaign about AI-assisted research, with a clear step for verifying generated claims.
This is a useful example of Dees weighing software choices by fit rather than assuming either an off-the-shelf app or custom development is always best.
Read more
Dees describes the “90/10 app problem”: a business may need only a small part of an app’s functionality for a specific workflow.
They contrast established tools such as Rewind and Webgility with small custom apps, arguing that AI has made tailored development more practical.
Custom development has a cost, they note, but in the right case it can mean less feature clutter and technical debt, with software shaped around how the business works.
For brands planning a campaign
A Shopify or business-software campaign could use this tradeoff to help merchants assess app sprawl, workflow needs, and the costs of a custom solution.
Open this for a broader view of attention as a way to prioritize business decisions, not simply a tactic for gaining visibility.
Read more
Dees distinguishes attention that matters from promotion that adds noise.
A spoofed e-commerce site and a client lead that could affect a team’s workload show how attention can shape operational decisions as well as marketing.
They apply the same idea to relationships: consider what deserves attention and why.
For brands planning a campaign
Could inform a campaign about prioritization, online trust, or customer relevance. The post points toward showing what merits attention rather than urging audiences to promote more.
This post shows how Dees condenses a longer conversation into a short list of actions and a question readers can answer.
Read more
After a Website Leader Podcast conversation with Richard Bliss, Dees distills the discussion into practical LinkedIn takeaways: post from a personal account, contribute meaningful comments, and make time for participation. They close by asking readers what they will start with, turning a recap into a prompt for action.
For brands planning a campaign
A useful model for podcast or professional-social campaigns that turn an interview into specific takeaways and invite audience participation.
Using AI to audit a company's credibility and positioning
Kevin Dees describes asking Google Gemini to research Robojuice and assess whether to hire the firm, then shares the three prompts used: one to audit credibility and contradictions, one to check market positioning against the industry vertical, and one to extract three pros and three cons.
Read more
The AI output describes Robojuice as a boutique consultancy with direct founder access, cites TypeRocket as an open-source framework behind more than 10,000 platforms, and flags limits around team scale, key-person dependency, and premium pricing.
Reusable prompt templates for auditing a brand or researching a vendor.
For brands planning a campaign
A brand evaluating AI-assisted process analysis gets a concrete example of turning a model into a structured vendor or self-audit, with the prompts and output format included.
Kevin Dees describes the 90/10 app problem: clients evaluate apps that cover 90% of features they do not need to get the 10% they do.
Read more
Rewind and Webgility are named as strong apps for complex needs such as backups and accounting.
The argument is that AI has made small custom Shopify apps more practical because the platform is well documented and models handle its APIs well.
The tradeoff is development cost versus less feature clutter and tech debt.
For brands planning a campaign
For a technology consultancy, this offers a decision framework brands can evaluate: when a client's specific operational needs justify custom development over a broad app, and what information belongs in that decision.
Attention as an operating decision, not a marketing metric
Kevin Dees argues attention belongs to business operations, not just marketing.
Read more
It goes to threats such as a spoofed e-commerce site, to client leads that change team workload, and to relationships that need you.
Promotion, socializing, and ads can add noise rather than cut through it.
Knowing when to pivot means weighing what has attention against what deserves it.
For brands planning a campaign
Useful for a brand that wants to discuss focus and prioritization rather than campaign tactics. The angle is operational attention, not reach or impressions.
In a Website Leader Podcast episode with Richard Bliss, Kevin Dees shares three points: LinkedIn prioritizes user activity over ads because subscriptions drive revenue; post from a personal account rather than a business page; and the algorithm amplifies conversations, so spend 30 minutes a day leaving meaningful comments.
Read more
The post frames this as guidance for leaders under pressure to keep up with a fast-moving digital landscape.
A daily commenting routine and posting from a personal account rather than a business page.
For brands planning a campaign
Brands can evaluate Kevin Dees for practical social-channel coaching, with a specific routine attached rather than general platform advice.
Kevin Dees praises Supermoon's recent updates, including Shopify integration, one-click Gmail integration, and AI auto-tagging controls, after talking with West Jones at Atlas Local.
Read more
They say they will check out Supermoon for Shopify and that they help Shopify clients with custom development while advising them on tools.
If paid-partnership status matters, ask whether this endorsement was sponsored.
For brands planning a campaign
Shows how Kevin Dees positions tool recommendations inside client work. A brand in the Shopify app or integration space can see what feature and integration news gets their attention.
Dees describes Robojuice as their own company and appears in its published positioning as a principal architect alongside Jeff Carver.
Read more
Their posts also show them working with Shopify as a development platform, evaluating the Shopify app Supermoon and hosting the Website Leader Podcast.
Robojuice
Employment
Dees calls Robojuice “our company” and shares a company audit that names them a principal architect alongside Jeff Carver, with clients working directly with the two principals.
Read more
The post describes Robojuice as a boutique consultancy working across Laravel, Shopify and WordPress.
The Robojuice audit credits the company's leadership with building TypeRocket, an open-source framework it says powers more than 10,000 professional platforms.
Read more
The post does not specify Dees's role; ask what they contributed to the project.
Dees builds custom Shopify apps for clients and advises them on tool choices.
Read more
They describe the “90/10 app problem” — clients evaluating apps that cover 90% they do not need in order to reach the 10% they do — and say AI has made small custom apps more practical than before.
Dees presents the Website Leader Podcast episode in which they talk with Richard Bliss about maximizing time on LinkedIn and shares three points from the conversation.
Kevin Dees writes about software decisions for real clients: which platform fits, when a small piece of custom code beats a broad app, and how AI changes both company research and what is worth building.
Read more
A good entry point is the 90/10 app problem post, where a client weighs apps that cover 90% of features they do not need to reach the 10% they do.
Next is the post describing a Gemini audit of their own firm, Robojuice, which publishes the three prompts used and treats the model's verdict as a draft to check rather than an answer.
Two further threads round out the picture: attention as an operating decision covering threats, client leads, and relationships, not just campaigns; and a Website Leader Podcast conversation with Richard Bliss on LinkedIn habits.
The tool and platform posts make the same practical case: judge software by fit, not feature count.
Windorpski uses the same first-person platform-choice approach as Dees: the post opens by rejecting the question clients actually ask and reframes it as a fit problem, arguing that most agencies push the platform they know rather than the one that suits the client.
Read more
They apply the same reasoning to Shopify apps. Both are agency operators explaining build decisions to business owners in plain language.
Windorpski focuses on conversion and traffic diagnosis, asking whether a site turns existing visitors into sales.
They frame platform selection for UK small businesses and ecommerce.
Dees focuses on engineering: the cost of custom development, the specifics of Shopify's API, and where AI fits into the research and build process.
Himanshu P. asks the same build-versus-buy question as Dees, opens with a similar reframe, and reaches the same conclusion: match the platform to the business instead of buying features you will not use.
Read more
The WordPress-versus-Shopify breakdown and their 90/10 post approach the same decision from different angles.
Himanshu P. argues with quantified scenarios, such as payment fees at volume and migration costs, and a four-quadrant matrix aimed at a broad technical and business audience. Dees starts with named client situations and specific tools, then extends the argument to what AI now makes practical to build rather than stopping at platform choice.
Sakas and Dees both write for business owners about adopting AI and the operational change it brings, and both publish frameworks rather than tips: Sakas's AI Adoption Continuum for agencies and Dees's prompt-based audit of a company.
Read more
Each examines a business's readiness, whether an agency's AI maturity or whether a vendor's positioning holds up.
Sakas writes about running an agency, covering AI adoption stages, when to sell, and hiring an operations lead.
Dees writes about the technology itself: what to build, what to buy, and how to use AI on the work.
Where Sakas places a business on a maturity curve, Dees offers prompts and describes the tradeoffs of a specific development decision.
The only measured audience figure in the posts here is account-level: 919 followers on Kevin Dees's LinkedIn profile, observed 8 October 2026.
Read more
The posts do not include platform analytics, a demographic breakdown or an audience survey, so they cannot show who follows them by geography, age, gender, employer, job function or seniority.
For a technology consultancy weighing Kevin Dees for guidance on applying AI tools to business process analysis, the following is small and its composition unmeasured.
The posts show the professional company they keep.
They engage publicly with Shopify tooling and agency peers—for example, commenting on Supermoon's Shopify, Gmail and AI auto-tagging features—and write for business owners and technical leads making software and tooling decisions.
This indicates the community they operate in, not who sees the posts.
Ask for recent post-level results—impressions, comments and profile visits—for posts on comparable topics (AI-assisted custom development, Shopify app decisions, LinkedIn practice) and formats, and whether any audience survey covers job function and seniority.
Until then, plan the placement as a small, unsegmented test rather than a scaled reach buy.
Kevin Dees's posts do not say they offer sponsorships or partnerships: no rate card, media kit, sponsored-format examples, or inquiry route for brand collaborations.
Read more
The posts do show advisory and delivery work.
In one, Dees describes running an AI-assisted audit of their own company, Robojuice, and lists the consultancy's strengths and constraints: direct access to principal architects, a small core team of under ten people, reliance on contractor networks for larger enterprise scopes, and premium mid-market pricing.
Another describes evaluating Shopify apps that cover only part of a client's workflow and building small custom apps instead; a third is a public endorsement of a Shopify tool.
A brand considering a paid collaboration should ask about sponsorship formats and terms.
A Shopify app vendor or development agency wants a creator to explain when custom, AI-assisted app development beats an off-the-shelf tool that only partly covers a merchant's workflow.
Reasons
The 90/10 app problem post makes the case: clients evaluate apps that cover 90% of features they do not need to get the 10% they do.
Kevin Dees names established tools — Rewind and Webgility — for complex needs such as backups and accounting, rather than dismissing off-the-shelf software outright. The recommendation reads as a judgment call, not a pitch.
The post states the tradeoff plainly: custom development still costs money, but the payoff is less feature clutter, less technical debt, and software built around how the business actually works. That suits a vendor willing to discuss cost against fit.
The AI audit post applies the same approach to research: three reusable prompts, a visible output, and a clear limit — a generated assessment starts the research rather than proving its own claims.
The topics and work sections show Kevin Dees writing for business owners and technical leads making software and tooling decisions — the audience this brief needs.
When to choose a specialist
Poor fit based on the information available.
Read more
Brief
A consumer brand wants a high-reach awareness push: mass impressions against a defined demographic, backed by a documented paid sponsorship package.
Reasons
The only measured audience figure is 919 LinkedIn followers, observed 8 October 2026. Without demographics, geography, job function, or seniority data, reach cannot be segmented or promised.
The posts provide no rate card, media kit, sponsored-format examples, or inquiry route for brand collaborations. Ask for these before considering a paid placement.
The content targets business owners and technical leads weighing software decisions, not a general consumer audience.
No post-level results — impressions, comments, or profile visits — were provided, so engagement among the small following is unmeasured.
Before treating this as more than a small, unsegmented test, ask for post-level results on comparable topics and an audience survey covering job function and seniority.
A worked campaign example
Angle
A live session could turn the audit prompts into a client-facing exercise: run the three-question audit on a volunteer company, then show where the model's output needs checking against public records and internal facts.
Brief
A technology consultancy wants Kevin Dees to help clients apply modern AI tools to business process analysis.
Steps
Open with the prompt set from the Robojuice post: one prompt to audit credibility and contradictions, one to test market positioning against the industry vertical, and one to extract three pros and three cons.
Run the prompts on a participating company and show the raw output, including claims about positioning, team scale, and pricing.
Walk through verification: which claims trace to public sources, which are inference, and which cannot be confirmed. Dees flags this limit in the post, so the session demonstrates it rather than merely asserting it.
Apply the 90/10 decision frame to the company's own tooling: where a broad app earns its place and where a narrow custom build fits better.
Give the audience the prompt templates and a one-page checklist for checking generated claims before acting on them.
Why it works
The session draws on two documented pieces of Dees's work — the AI audit prompts and the 90/10 app argument — and pairs the demonstration with the verification step they raise. The consultancy leaves with a repeatable method and a checking routine, not general commentary about AI.
Choose Sakas when the consultancy needs to judge where a client's business actually sits in AI adoption before recommending tools.
Read more
A post lays out five named levels: no meaningful use; incidental use that arrives inside existing software such as Google Workspace, Canva, Adobe, Microsoft, Notion or HubSpot; intentional use of ChatGPT, Claude or Gemini with shared prompts and department templates; operational use, where AI moves into workflows with n8n, Make or Zapier and agents acting inside project or CRM tools; and integrated use, where systems pull on internal data through RAG or embeddings.
They place most agencies between levels two and three and say Sakas & Company is moving into level four.
The post ties adoption to how an organization handles change, treating an agency's AI competence as a proxy for adaptability and operational maturity.
That is a readiness assessment a consultancy can sell to a client worried about whether an AI rollout will take, and a framework Dees does not offer.
Sakas writes for agency owners, and the argument runs toward ownership decisions: the post says that if championing AI adoption feels draining rather than energizing, it may be time to explore selling the agency.
That differs from the process-analysis work a consultancy may need.
The consultancy gets an adoption scorecard and a change-readiness judgment, not the applied method Dees demonstrates.
The posts do not show an AI tool run over a specific company's public record, the output inspected, or the claims checked; ask for a worked example.
Another post recommends an operations candidate, useful as an example of operational leadership but not AI process analysis.
Choose this alternative when the question under the AI request is really build versus buy for one capability.
Read more
A post reframes the familiar custom-or-platform choice as a matrix of differentiation value against volume: buy commodity services such as accounting, HR and project management; use a platform with custom work where volume is high and commodity services still carry cost, illustrated with $50M in annual processing and $1.45M a year in Stripe fees; build only the differentiating slice that makes a business unique; or go fully custom where no platform can scale with the need, as the post says of Netflix.
It also lists costs most buyers leave out—platform lock-in and migration, technical debt from small tweaks, constraint creep and opportunity cost—and a working rule: buy commodity services, customize mid-tier differentiators, build core IP.
A consultancy that has to place an AI tooling decision on that same map and argue it with numbers in front of a client gets a repeatable structure here.
Dees's posts focus on WordPress against Shopify, custom software against platforms, and the cost of flexibility—not AI tools or business process analysis.
Applying them to a campaign about AI and business process analysis means the consultancy must do the translation itself.
Ask for a prompt method, an audit example and a verification step for generated claims.
Use this as framing around the AI question, not as an answer to it.
Choose Windorpski only if the consultancy's client work is web and e-commerce delivery and the client's need shifts from process analysis to why a site is losing the sales its traffic should already produce.
Read more
A post separates search visibility from conversion: SEO can be working while the landing page makes visitors hunt for answers, the mobile experience adds friction or the next step is unclear.
It turns that into three checks to run before increasing a marketing budget.
The WordPress-versus-Shopify post takes the same client-first stance, opening by saying the weekly platform question is the wrong one and replacing it with what the business must be able to do in eighteen months, while conceding that most agencies push the platform they are more comfortable with.
Dees's posts cover websites, SEO and platform selection; they say nothing about AI tools, process analysis or organizational diagnostics. This alternative addresses a different problem and would leave a need to apply AI to business process analysis untouched.
Five LinkedIn posts by Kevin Dees cover AI-assisted company audits, custom Shopify development, business attention, LinkedIn habits, and a Shopify tool update.
The posts do not say when they were published; the account was observed on 8 October 2026, and the wider section spans September 2023 to September 2026.
The similar-creator comparisons use two LinkedIn posts each from Michal Windorpski, HIMANSHU P., and Karl Sakas.
No video, transcripts, post-level performance data, audience demographics, or sponsorship terms are published here, so reach and paid-partnership capability remain open questions rather than documented facts.
