Tom Shivers

Several problems reinforce one another: weak rankings, technical issues that suppress visibility, content that misses high-intent demand, thin conversion paths, and paid traffic carrying more of the load than it should.

• Because these problems interact, the order of repairs matters.
• Fixing the right things in the right order unlocks the channel; fixing the loudest thing first does not.

A mid-sized B2B ecommerce company was stuck at $3.1M in annual revenue, spending heavily on high-CPC Google Ads keywords as organic performance plateaued. After addressing the diagnosed problems in sequence, Tom Shivers reports a 54% lift in organic conversions, a 33% increase in average order value, and 140% ecommerce revenue growth, taking revenue from $3.1M to $7.5M.

Read original

The post argues that symptoms vary by industry, but the underlying mechanics do not, so the same diagnosis keeps applying. It promises to break down the blockers one at a time, starting with SEO misdiagnosis.

The resource-imbalance blocker is explained as a mechanism, not a slogan: chasing a hyped channel shifts budget and attention away from a working part of the system, degrading it while the team figures out the new channel.

The post takes the diagnostic pattern from a single-company case and turns it into a named, ordered list of causes Shivers says recur across sectors.

Read original

Inside ChatGPT, ACP enables AI-driven product discovery and comparison, conversational shopping, cart building, checkout through delegated payments, and post-purchase handling by the retailer.

• The retailer remains the merchant of record, answering the post's central question: how can an AI agent shop without becoming a marketplace?
• The open-source protocol is designed to be interoperable, allowing any agent, payment processor, or retailer to adopt it.
• ACP extends an existing ecommerce stack to AI agents rather than replacing it.

The adoption list is specific: provide a structured product feed to OpenAI, implement ACP checkout endpoints, support delegated payments, keep pricing, inventory, and fulfillment logic clean and consistent, and retain control over merchandising, policies, and customer communication.

Read original

Unlike marketplaces such as Amazon or Walmart, the merchant retains the transaction, post-purchase experience, and customer relationship.

• The post reframes the buyer path in three lines: the structured feed is the new homepage, the product graph is the new search results page, and the AI agent is the new buyer journey.
• Incomplete, inconsistent, or poorly structured data means a product will not be recommended.
• Merchant Center Next feeds, local inventory feeds, and structured data determine whether a retailer appears in AI Mode or qualifies for agent-led checkout.
• The four recommended actions are to fix product feeds, strengthen the brand entity with schema, publish authoritative content AI can cite and track citation share, and make payment, fulfillment, and policies machine-readable.
• The post offers a feed-readiness checklist for auditing ecommerce brands.

The post leads with a three-part reframing: when a retailer asks where its storefront went, the answer is that the product feed now does the job the homepage used to do.

While the ACP post explains what agents can already do, this post frames the same shift as an urgent data-preparation project and offers a short list of fixes.

Read original

The approach favors practical resources over opinion alone.

A post announces a free eCommerce ROI forecast calculator that estimates how much revenue a business could gain by optimizing its strategy.

• Readers enter a few inputs for a personalized ROI projection, then are invited to try the calculator and share results in the comments.
• To assess the calculator, ask for its inputs, method, and output format.

Read original

It explains what ACP is, who built it, and why it arrived before Google's competing standard.

Tom Shivers explains that the Agentic Commerce Protocol (ACP), built by OpenAI and Stripe, has quietly connected retailers and AI agents for months, predating Google's Universal Commerce Protocol.

• The post walks through what ACP enables inside ChatGPT: AI-driven product discovery and comparison, conversational shopping, cart building, checkout through delegated payments, and retailer-managed fulfillment and post-purchase service.
• The retailer remains the merchant of record, so the AI does not become a marketplace.
• Shivers notes that the protocol is open-source and designed to be interoperable, allowing any agent, payment processor, or retailer to adopt it.
• The practical section lists what retailers must supply: a structured product feed to OpenAI, ACP checkout endpoints, delegated payment support, clean pricing and inventory data, and continued control over merchandising and customer communication.
• ACP extends an existing ecommerce stack into AI agents rather than replacing it.

For brands planning a campaign

Useful for a sponsor selling ecommerce infrastructure and looking to show retailers how to appear in AI shopping assistants through an explainer, webinar, or article.

Tom Shivers describes Google's Universal Commerce Protocol as a standard that lets AI agents compare products, check inventory, build carts, and complete checkout without visiting a retailer's website.

• Unlike Amazon or Walmart, UCP keeps the customer relationship, transaction, and post-purchase experience with the merchant.
• Shivers says this reshapes the path from discovery to purchase: the structured feed becomes the new homepage, the product graph becomes the new search results page, and the AI agent becomes the new buyer journey.
• Incomplete, inconsistent, or poorly structured data can keep a product from being recommended.
• Merchant Center Next feeds, local inventory feeds, and schema determine whether a retailer appears in AI Mode or qualifies for agent-led checkout.
• Shivers lists four actions: fix product feeds, strengthen the brand entity and schema, publish authoritative content that AI can cite, and make payment, fulfillment, and policies machine-readable through Merchant Center Next.
• The post offers a feed-readiness checklist for auditing ecommerce brands.

For brands planning a campaign

A natural hook for a sponsored checklist, audit offer, or article aimed at ecommerce brands preparing product data for AI-led sales.

Tom Shivers describes a mid-sized B2B ecommerce company stuck at $3.1M in annual revenue, carrying most of its traffic through costly high-intent Google Ads keywords while organic performance had plateaued.

• The diagnosis listed stalled organic rankings, technical SEO problems limiting visibility, content that did not match high-intent demand, weak conversion paths, and paid traffic carrying too much of the load.
• Shivers says fixing the right problems in the right order unlocked the channel, and reports a 54% lift in organic conversions, a 33% increase in average order value, and 140% ecommerce revenue growth, taking the business from $3.1M to $7.5M.
• The closing point is that growth usually stalls from a cluster of misdiagnosed problems rather than one large issue.

For brands planning a campaign

A centerpiece case study for a sponsor seeking proof that SEO and conversion work can turn around a B2B ecommerce account.

Tom Shivers says the same five blockers appear across B2B, B2C, SaaS, ecommerce, automotive, healthcare, home services, and food and beverage: misdiagnosed SEO issues; channel hype that creates a resource imbalance and starves working parts of the system; wasted ad spend; content that does not match intent; and teams stretched too thin to fix root causes.

• The post argues that symptoms vary by industry but the underlying mechanics do not.
• It says each blocker will be broken down in turn, starting with the most common SEO misdiagnosis.

For brands planning a campaign

A concise statement of themes a sponsor can build a series around, especially for marketers comparing channel investments.

Tom Shivers announces a free ecommerce ROI forecast calculator that estimates how much revenue a business could gain by optimizing its strategy. A few inputs produce a personalized ROI projection; the post invites people to try it and share their results in the comments.

For brands planning a campaign

A template for a co-branded calculator, assessment, or interactive resource that gives a sponsor a lead-generating asset rather than a written piece.

The first covers OpenAI and Stripe's Agentic Commerce Protocol (ACP), which they describe as arriving before Google's Universal Commerce Protocol and as the bridge that let ChatGPT compare products, build carts, and complete checkout through delegated payments while the retailer stays merchant of record.

• The second covers Google's Universal Commerce Protocol (UCP), which lets agents compare products, check inventory, build carts, and finish checkout without visiting the retailer's site, while the merchant keeps the transaction, post-purchase experience, and customer relationship.
• Both posts frame the shift in the same terms: the structured feed replaces the homepage, the product graph replaces the search results page, and the AI agent becomes the buyer journey.

Each post sets up a contrast—ACP versus UCP, or agent-led checkout versus marketplace—and lists concrete retailer actions.

For brands planning a campaign

A B2B ecommerce software company can sponsor a side-by-side explainer or webinar showing buyers what ACP and UCP each require and where they overlap. The posts offer a ready structure: protocol origin, what the agent can do, who owns the customer, and retailer setup steps.

Examples: Source 1 · Source 2

The steps include providing a structured product feed, implementing checkout endpoints, supporting delegated payments, and keeping pricing, inventory, and fulfillment logic consistent.

• The UCP post adds strengthening the brand entity with schema, publishing authoritative content that AI can cite, tracking citation share, and making payment, fulfillment, and policies machine-readable through Merchant Center Next.
• Tom Shivers states the consequence directly: incomplete or inconsistent data means the product will not be recommended.

Turns a platform change into an ordered preparation list for retailers.

For brands planning a campaign

This is the most sponsor-ready topic for a feed-readiness or data-quality product. A brand can fund a checklist, audit template, or working session built around these steps, using the checklist as a lead magnet and the software as a follow-on.

Examples: Source 1 · Source 2

The diagnosis lists stalled organic rankings, technical SEO problems suppressing visibility, content that did not match high-intent demand, weak conversion paths, and paid traffic carrying too much of the load. After the problems were fixed in what Tom Shivers calls the right order, they report a 54% lift in organic conversions, a 33% increase in average order value, 140% ecommerce revenue growth, and revenue moving from $3.1M to $7.5M.

Sequence matters: diagnose a cluster of problems, fix them in order, then report channel and revenue results together.

For brands planning a campaign

Suitable for a sponsored case study, teardown webinar, or diagnostic workshop aimed at B2B ecommerce teams. The before-and-after numbers and ordered repair list give a brand concrete hooks; confirm the underlying data before reusing the figures.

Examples: Source 1

The argument is that symptoms differ by industry but the underlying mechanics do not. The post announces that each blocker will be broken down separately, starting with SEO misdiagnosis.

Generalizes from a list of industries to a short, named set of causes, then commits to a sequential series.

For brands planning a campaign

A foundation for a sponsored series or one-per-blocker content program rather than a single asset. A platform selling SEO, paid media, or content tooling can attach to the specific blocker it addresses.

Examples: Source 1

The post is short and promotional: it states that businesses lack tools to forecast ROI, offers the calculator, and asks readers to share results in the comments.

Ask for the calculator's inputs, methodology, and output format before assessing its capabilities.

For brands planning a campaign

Evidence that Tom Shivers builds and promotes practical, interactive lead tools, supporting a sponsored calculator, benchmark, or forecast tool as a collaboration format. Test the existing calculator before proposing a replacement or co-branded version.

Examples: Source 1

Both treat agentic commerce as a readiness problem with concrete gaps to close, not a forecast.

The posts come from inside a commerce platform vendor and rely on its own scans and launch announcements, so the argument runs through a product being sold. Tom Shivers explains open protocols and retailer preparation without a platform behind the explanation, keeping agent-side mechanics separate from any offering.

Read original · Original 2

The focus is narrower and storefront-level — Shopify theme data, Meta tracking quality, consent and event matching for DTC brands — while Tom Shivers works at the protocol layer and diagnoses stalled growth across B2B, SaaS, and services businesses.

Read original · Original 2

It covers the same subject and aims to give readers a map rather than a headline.

The treatment is trend-led and consumer-facing, opening with shoppers asking a chatbot for product advice and walking through payment standards from Google, Mastercard, Visa, and PayPal. Tom Shivers stays closer to two named protocols and the retailer setup steps behind them, while adding ecommerce growth diagnosis that has no counterpart in these posts.

Read original · Original 2

The two posts are vendor launch and joint-research announcements: Mirakl Agentic Activation, with Agentic Product Enrichment and Agentic Channels connecting merchants to large language model platforms starting with Microsoft Copilot, and a Mirakl and Merkle readiness playbook presented with market figures — fewer than 1% of scanned product pages scoring above the threshold where AI agents reliably surface products, $67 billion of Cyber Week 2025 sales influenced by AI agents, 393% year-over-year growth in AI-referred traffic to U.S. retail sites, and a projection that 55% of U.S. ecommerce spending will soon be agent-assisted. That suits a sponsor that wants benchmark numbers and a marketplace-infrastructure framing, and it shows a comparator already speaking to enterprise buyers through a named research partnership.

The posts promote an employer's product launch and argue that businesses need to get ready; they don't explain how the Agentic Commerce Protocol or the Universal Commerce Protocol works.

• They offer no step-by-step implementation sequence, ordered fix list or before-and-after campaign result.
• Sponsors seeking an independent explainer of how the protocols work, what a retailer must supply and in what order should ask Tom Shivers for more measured detail.
• Those looking for vendor-style market framing and launch news will find that here, in exchange for less instructional depth.

Examples: Source 1 · Source 2

The posts are concrete about platform changes and what to do about them.

• One audits what businesses send to Meta as the company broadens use of shared activity beyond ads, listing five checks before a July rollout: pixel and Conversions API deduplication, purchase values and currencies matching store revenue, event coverage at the right stage, consistent event matching parameters, and documented consent and governance.
• The other works through Shopify's May 2026 announcement that merchants can customise /agents.md, /llms.txt and /llms-full.txt, then lists what to audit so AI systems can understand a store: product titles, collection structure, product page copy, schema and structured data, Merchant Center feeds, reviews and product context, shipping and return clarity, and AI-readable theme data.
• If the sponsor's readers run Shopify or Meta and want configuration-level steps, this is closer to their working day than a protocol overview.

The scope is narrower: tracking quality and storefront readability on two specific platforms.

• The posts don't cover delegated payments or checkout endpoints, or compare the two agentic commerce standards—topics central to a sponsor's standards explainer.
• They react to platform announcements, offering audit lists rather than a demonstrated method and measured outcome.
• Sponsors needing a retailer data-preparation spine across feeds, payments, fulfilment and policies should ask Tom Shivers for more breadth; those seeking checklist content for a named platform can trade breadth for specificity.

Examples: Source 1 · Source 2

One post maps four agentic payment protocols side by side: Google's Agent Payments Protocol, Mastercard's Agent Pay, Visa's Trusted Agent Protocol and PayPal's Agentic Commerce Platform, including which card networks and payment companies back each.

• The other traces agentic commerce through eBay, Amazon, social commerce and short video, then uses the 'ChatGPT GlowUp' example of shoppers asking a chatbot for skincare advice.
• It also notes that traffic from answer engines is already around 10% for some sites.
• Together, the posts offer a useful spine for a webinar on how agent-driven transactions get authorised, who is competing for the payment rail and how shopping behaviour is changing.

The posts list protocols and trends rather than show how a retailer implements them.

• They don't detail product feeds, schema, Merchant Center or checkout endpoints, or provide a before-and-after result for a sponsor's audit segment.
• The consumer-trend framing also sits further from a B2B ecommerce practitioner audience than Tom Shivers's two protocol explainers and promised feed-readiness checklist.
• Sponsors deciding what a brand's data must look like before an agent will recommend it should ask for that implementation detail; those focused on payments and shopper behaviour can trade implementation depth for a wider, more consumer-facing view.

Examples: Source 1 · Source 2