Micky Weis

A better starting point is to find real problems worth solving.

• The next step depends less on which model you pick than on what data you can access and the governance around it.
• Weis recommends connecting business tools, using read-only access while learning, scheduling useful reports, and feeding conversations, emails and support tickets into the model to find processes worth improving.
• Their working habits include one folder per client, a file describing the client from A to Z, brand DNA and previous ads in an assets folder, and md-files and skill files inside Claude.

Micky Weis lists a current stack of Claude, ChatGPT, Whisper Flow and BigQuery, with connectors for Shopify, Figma, Canva, Meta and Drive.

• With Shopify, Klaviyo and Meta connected, a brand can produce sharp performance reports, such as a scheduled morning report.
• Start with easy wins rather than a big-scale rollout.

Read original

The approach covers email and customer dialogue first, then branding, content, onsite work and data.

• Check email flows, discount codes, calls to action and triggers, and build lists further with lead ads and onsite sign-up work.
• Think about branding earlier: the better known a brand is before Q4, the cheaper performance becomes later.
• Plan content in advance.
• Onsite, keep category pages in order, move sold-out products, build upsells on product pages and in the cart, and test return windows.
• For data work, connect data to an AI model and ask it to review previous Black Friday and Christmas sales.

Micky Weis suggests checking whether discount codes have leaked into coupon-code apps, using a content structure in a tool such as Focal, considering an MCP connector in Claude that scrapes Meta's ad library for inspiration, testing whether 365-day returns shift conversion, and opening your own site in an incognito window to count how many pop-ups irritate visitors.

Read original

Without knowing when third-party dealers ran discounts, month-to-month revenue comparisons can mislead: a bad March may simply be a month without discount noise.

• Discount-code apps complicate the picture because codes spread quickly through apps that collect and share them.
• A supposedly exclusive newsletter code can become public within days, and the real reach of the discount is rarely what the brand thinks.
• Pull your own discount days, dealer discount days and code-use counts into one view and compare them with sales over time.

Micky Weis built a small dashboard in Lovable that combines their own discount days with dealers' discount days, so they can see which days discount codes correlate with extra sales and which months they do nothing. Dealer data is still entered manually because some dealers run older systems.

Read original

Periods with more video, talks and posts produce more enquiries.

• The enquiries have changed: in 2018 they were mostly about Meta advertising, e-commerce, SEO and a little content; today they involve larger conversations about AI, scalability, internationalisation and finding the right partners across tasks.
• Weis also notes a broad digital shift in 2020–2022 and a withdrawal in 2021–2022 because of stress.
• They express pride in passing work on to other agencies and people.

WeisOS includes recurring data from the channels Micky Weis uses, so activity can be compared with enquiry volume and topic. The post names a long list of people and businesses they have referred work to, and notes that people help each other in a small country like Denmark.

The enquiry mix shifted from channel-specific work such as Meta advertising, e-commerce and SEO in 2018 toward AI, scalability, internationalisation and partner selection in later posts.

Read original

This makes Weis a route to recommendations and referrals in that area rather than necessarily the hands-on Klaviyo or Shopify implementer. The same priority appears in the WeisOS post, where they say they are proud of work sent on to other agencies, people and businesses.

The Klaviyo and Shopify post is a direct request for freelancer names and references. In the WeisOS post, Micky Weis lists numerous people and businesses they have passed enquiries to and thanks them.

Read original · Original 2

The post shows how they think about AI in e-commerce and what they use day to day.

In response to a recurring inbox question, Micky Weis advises solving real problems rather than chasing every tool.

• They list their own stack — Claude, ChatGPT, Whisper Flow and BigQuery for data — plus connectors set up in Claude for Shopify, Figma, Canva, Meta and Drive.
• Their point is that the next step with AI depends on the data you can access and the governance where you work: connect Shopify, Klaviyo and Meta, for example, and you can produce sharp performance reports, including a scheduled morning report.
• They also describe their workflow: one folder per client, with a file describing the client from A to Z, plus brand DNA, previous ads and newsletters in an assets folder, and md-files and skill files inside Claude.
• They mention having taught teachers at Erhvervsakademi Kolding and say they are not an expert, just using AI where it solves something.

For brands planning a campaign

For a Danish Shopify brand, this is the basis of a data-and-AI session: pick one recurring report the team already writes by hand, connect the underlying sources with read-only access, and automate that report first. The client-folder structure also offers a practical way to give an outside partner brand context without unrestricted account access.

This concrete pre-season checklist shows the four areas Micky Weis prioritises and the specific checks they run, making it easy to compare their approach with your own Q4 plan.

Micky Weis’s Q4 checklist covers four areas.

• For email and customer dialogue, they recommend reviewing every flow: check whether discount codes are still the right ones, whether they have ended up in coupon-code apps, whether calls to action land, and whether triggers make sense.
• Then keep building the list with lead ads and onsite sign-up work.
• On branding, their argument is that the better known a brand is before Q4, the cheaper performance becomes later; plan top-funnel spend earlier and lower-funnel spend later.
• On content, they say they started too late themselves and got the best results by setting a structure in advance.
• They cite Focal as an example and suggest an MCP connector in Claude that scrapes Meta’s ad library for inspiration.
• Their onsite advice, drawn from multi-brand setups, is to keep category pages in good shape, move products as they sell out, build upsells on the product page and in the cart to lower customer acquisition cost and raise order value, and test the returns window: 365 days can shift conversion more than people expect.
• For data, they suggest connecting it to an AI model and asking it to analyse previous Black Friday and Christmas sales, with Klaviyo and Meta in the same setup.
• Their final practical check: open your site in an incognito window and count your own pop-ups.

For brands planning a campaign

This offers a ready-made brief agenda: a flow and coupon-code audit, an earlier branding-budget case for management, a content structure with a set production tool, and an onsite/returns test. A brand could ask Micky Weis to run the audit first, then use the findings to choose one change to test before Q4.

It’s a small, complete example of building the missing answer yourself, and a useful way to assess whether their recommendations are grounded in measurement or habit.

Most people Micky Weis talks to can say whether a month was good or bad, but few can say how many orders were sold with a discount across channels.

• They built the answer themselves: a small dashboard in Lovable that pulls their own discount days and their dealers’ discount days into one picture, showing which days discount codes correlate with extra sales and which months they do nothing.
• With third-party dealers, one month can look like a discount every third day and the next month none at all.
• Comparing revenue month to month without knowing those days is comparing apples and pears: a bad March may simply be a month without discount noise.
• They also describe coupon-code apps as a problem: run many codes on a Shopify shop and they spread quickly through apps that collect and share codes, so an “exclusive” newsletter code is public within days and the real reach of the discount is rarely what the brand thinks.
• Dealer data is still entered manually because those dealers run older systems.

For brands planning a campaign

A Shopify brand with dealers or resellers could commission the same picture: one dashboard combining its own discount days, dealer discount days and code usage, so discount decisions are based on measured incremental sales. It would also show how exposed the brand’s codes are in coupon-code apps before the next campaign.

The post explains why they often pass enquiries to other agencies and freelancers rather than taking everything on themselves.

Micky Weis describes a self-built system, WeisOS, and one of the things they put into it: enquiries, with exports running in a loop from the channels they use.

• When they push harder with video, talks and posts, more enquiries come in.
• Their activity has varied over the years: they were very active in 2018, when competition was slightly lower and AI was not part of the picture.
• The enquiries have changed too.
• In 2018, they were mostly about Meta advertising (then Facebook), e-commerce and SEO, with a little content.
• Today, they involve larger conversations about AI, scalability, internationalisation and finding the right partners across tasks.
• They also mention stepping back in 2021–2022 because of stress.
• They name a long list of people and businesses they have passed work on to and thank them, noting that there is something unique about a small country like Denmark where people help each other.

For brands planning a campaign

Micky Weis describes a one-person operation that refers work out.

• Before briefing them, ask what they will do personally and who else may be brought in.
• Their account also suggests that briefs covering AI, scalability and internationalisation are common enquiries for them now.

Their current stack includes Claude, ChatGPT, Whisper Flow and BigQuery, with connectors to Shopify, Figma, Canva, Meta and Drive.

• They stress that data access and governance determine what is possible, recommend read-only database access while learning, and suggest scheduled performance reports.
• They also describe pulling Slack conversations, emails and support tickets into Claude to find processes worth improving, starting with small wins rather than a big-scale rollout.
• For client work, they keep per-client folders with a full description, brand DNA, previous ads and newsletters.

Start with data availability and governance; use read-only access; connect business tools; build client context files; schedule reports; mine support, email and chat data for process improvements; take easy wins first.

For brands planning a campaign

A Danish Shopify brand can evaluate Micky Weis for an AI and data diagnostic focused on practical data access and small operational wins, not an immediate platform overhaul. The brief should name the problem to solve, the data sources to connect and the first read-only reports or process fixes to test.

Examples: Source 1

They start with email and customer dialogue: review all flows, check whether discount codes are still correct and whether they have leaked into coupon apps, assess calls to action and triggers, and build lists with lead ads.

• They argue for earlier top-funnel branding so lower-funnel performance gets cheaper later, and for structuring content production well ahead of Q4, including using Claude connectors to scan Meta's ad library for inspiration.
• Onsite, they suggest category page discipline, moving sold-out products, upsell on product and cart, and testing longer return windows.
• They also connect Klaviyo and Meta data with AI to review past Black Friday and Christmas sales, and check the site in incognito for popup overload.

Audit flows, codes, calls to action and triggers; build top-funnel brand earlier and lower-funnel conversion later; set a content structure; use AI to inspect Meta's ad library and prior holiday sales; test onsite and return-window changes.

Micky Weis also says they often start content too late themselves, so the sequence is advice rather than a record of flawless execution.

For brands planning a campaign

This maps well to a Q4 planning brief: use it as a checklist for email flows, brand timing, content pipeline, onsite merchandising and past-season data analysis. Ask Micky Weis to turn the checklist into a priority order and assign owners rather than treating every item as equally urgent.

Examples: Source 1

The goal is to see when codes correlate with incremental sales and when they do not.

• They warn that dealer promotions can make one month look like heavy discounting and the next like none, so month-to-month revenue comparisons without that context can mislead.
• They also say discount-code apps can quickly expose supposedly exclusive newsletter codes, changing a discount's real reach and effect.
• Dealer data may still need to be entered manually because some partners run older systems, but the consolidated view makes the discount effect visible over time.

Pull own discount days, dealer discount days and code-use counts into one dashboard, then compare those patterns with sales over time.

This is based on Micky Weis's own dashboard and observations, not a published client case study with independently verified results.

For brands planning a campaign

To assess Micky Weis for promotion planning, ask whether they can consolidate own and dealer discount activity before measuring lift. The plan should include access to promo calendars, code usage data and dealer promotion dates, plus a check for code leakage.

Examples: Source 1

The post positions them as a source of recommendations and referrals, not necessarily as the hands-on Klaviyo or Shopify implementer.

The demand is explicitly for freelancers, not a stated offer from Micky Weis to execute the work.

For brands planning a campaign

If the need is detailed Klaviyo and Shopify build work, ask Micky Weis for the named freelancers and references, then scope the implementation with those specialists directly. Confirm which work they will handle and which they will refer out.

Examples: Source 1

They observe that periods with more video, talks and posts produce more inquiries.

• In 2018, inquiries were more targeted around Meta advertising, e-commerce, SEO and some content.
• Today, inquiries are more complex, covering AI, scalability, internationalization and finding the right partners across tasks.
• They also note a broad digital shift in 2020–2022 and a stress-related withdrawal in 2021–2022, and express pride in referring work to other agencies and people.

For brands planning a campaign

A brand evaluating Micky Weis should expect conversations that extend beyond channel execution into AI, scaling, internationalization and partner selection. If the need is narrower, ask which parts they would own and which they would refer out.

Examples: Source 1

They say one addition is enquiry tracking with exports from their channels.

View example

Ask whether the engagement was employment or a contract.

View example

Christian Lunøe Gøtzsche takes the practical approach further, laying out a concrete agent workflow for a 50M EUR B2B company entering the German market, naming each agent's output, and arguing that governance and compliance belong inside the workflow, not bolted on afterwards.

Christian Lunøe Gøtzsche writes from an executive and commercial-leadership vantage point about agent architecture, EU AI Act and GDPR guardrails, and organisational adoption. Micky Weis focuses on day-to-day e-commerce operations — data access, Shopify and Klaviyo connections, Q4 planning and discount measurement — and is more likely to hand a technical build to a named specialist.

Read original · Original 2

John Nielsen's walkthrough of entering a client's building and plugging into a meeting-room socket uses the same teaching method as Micky Weis's incognito-window pop-up test: start with a concrete episode, then give the steps.

John Nielsen focuses on cybersecurity and digital sovereignty — red teaming, physical access control, and mail authentication across hundreds of Danish domains. Micky Weis focuses on growth and commerce: data plumbing, Q4 trading and whether promotions pay off.

Read original · Original 2

Stefan Kløvning's no-code, low-code and full-code test — simple and predictable workflow, more rules but still visual, complex or business-critical — reflects the same instinct to offer a practical rule of thumb that Micky Weis applies in recommending easy wins and read-only data access.

Stefan Kløvning writes about system integration and software reliability, including the risk of relying on a single third-party provider such as Cloudflare. Micky Weis writes about how brands should use the systems they already have, and the commercial decisions — promotions, Q4 sequencing, enquiries — those systems are meant to serve.

Read original · Original 2

A post describes an agent sequence for a B2B company entering Germany, with a compliance check before data reaches the CRM. That is a closer match than Micky Weis when the priority is structured agent architecture and governance rather than e-commerce reporting and seasonal merchandising.

The post presents an example, not a delivered Shopify, Klaviyo or Meta reporting system or Q4 e-commerce results. It focuses on B2B expansion, so ask how the approach would translate to your data and channels.

Examples: Source 1 · Source 2

A post walks through an authorized physical-access test, explains how penetration testing differs from red teaming, and gives steps such as checking network ports and asking who receives alerts. That hands-on security focus is more useful than Micky Weis’s e-commerce and AI planning when the brand’s urgent question is whether its people, premises or network controls would detect an intrusion.

The reviewed examples concern cybersecurity, not e-commerce performance, AI data analysis or Q4 preparation, so they don’t support choosing John Nielsen for a Shopify data setup or seasonal plan.

Examples: Source 1 · Source 2

One post gives a practical way to choose between no-code, low-code and custom integrations, based on workflow complexity and maintenance needs.

• Another uses a cloud-service outage to argue for dependency mapping, backups and contingency plans.
• Those examples are more relevant than Micky Weis’s work when the campaign hinges on integration design or operational continuity rather than marketing analysis.

The posts support a systems-integration and resilience comparison, not a claim that Stefan Kløvning can deliver Micky Weis’s proposed marketing work. Ask for an example of an e-commerce Q4 plan, discount analysis or Shopify–Klaviyo–Meta reporting.

Examples: Source 1 · Source 2