<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Brand Design Ltd.</title>
    <description>The latest articles on DEV Community by Brand Design Ltd. (@branddesignltd).</description>
    <link>https://dev.to/branddesignltd</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4152559%2F2718834d-d08f-41c7-badf-87b0bd0d763f.png</url>
      <title>DEV Community: Brand Design Ltd.</title>
      <link>https://dev.to/branddesignltd</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/branddesignltd"/>
    <language>en</language>
    <item>
      <title>When AI leaves your brand out: connect answer evidence to a site-footprint audit</title>
      <dc:creator>Brand Design Ltd.</dc:creator>
      <pubDate>Thu, 19 Nov 2026 00:28:17 +0000</pubDate>
      <link>https://dev.to/branddesignltd/when-ai-leaves-your-brand-out-connect-answer-evidence-to-a-site-footprint-audit-5di5</link>
      <guid>https://dev.to/branddesignltd/when-ai-leaves-your-brand-out-connect-answer-evidence-to-a-site-footprint-audit-5di5</guid>
      <description>&lt;p&gt;Disclosure: B2A Radar is a product of Brand Design Ltd., the publisher of this article. This article was prepared with AI assistance and checked against the product's public pages and implementation. The examples below are illustrative, not customer results.&lt;/p&gt;

&lt;p&gt;A visibility report becomes useful when it helps a team decide what to investigate next. Knowing that an AI answer omitted your business is the starting point. The next step is to inspect the answer, its sources and the information your website makes available.&lt;/p&gt;

&lt;p&gt;We built B2A Radar to bring those steps into one workflow: recurring measurements of AI answers, followed by a site and brand-footprint diagnostic that gives the team concrete findings to act on.&lt;/p&gt;

&lt;h2&gt;
  
  
  Start with the buying question
&lt;/h2&gt;

&lt;p&gt;An evaluation needs a defined context. A question about a local service in Bulgarian and a question about an international supplier in English do not test the same buying situation.&lt;/p&gt;

&lt;p&gt;In Radar, questions have a market and language. Checks cover GPT, Claude, Gemini and Perplexity. The report keeps three observations distinct:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;A &lt;strong&gt;mention&lt;/strong&gt; records that the answer names the brand.&lt;/li&gt;
&lt;li&gt;A &lt;strong&gt;recommendation&lt;/strong&gt; records that the answer actually suggests it for the buyer's need.&lt;/li&gt;
&lt;li&gt;A &lt;strong&gt;citation&lt;/strong&gt; identifies a source page linked in the answer.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;You can inspect the original wording, the brands appearing alongside yours, and the cited pages. This makes the result reviewable. An aggregate score is more useful when the evidence behind it remains available.&lt;/p&gt;

&lt;h2&gt;
  
  
  Inspect the site's footprint
&lt;/h2&gt;

&lt;p&gt;The diagnostic examines five areas:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Area&lt;/th&gt;
&lt;th&gt;What the team investigates&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Access for AI systems&lt;/td&gt;
&lt;td&gt;Whether the checked pages can be reached and their content read.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Structured data&lt;/td&gt;
&lt;td&gt;Whether the machine-readable description of the business, services and products is present and consistent with the content.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Content readability&lt;/td&gt;
&lt;td&gt;Whether the pages explain what the business does, what it offers and how a customer proceeds.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Brand-information consistency&lt;/td&gt;
&lt;td&gt;Whether names, contact information and profile references tell a coherent story.&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;B2A readiness&lt;/td&gt;
&lt;td&gt;Whether the site describes its offers and the route by which a customer or agent can act on them.&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;The useful output is the finding and the proposed action. A low score alone does not tell a developer which page needs attention or what needs to change.&lt;/p&gt;

&lt;p&gt;Consider an illustrative case: a supplier is absent from an answer about a particular service. The site check finds that the relevant service page is unavailable to the checker. The original answer cites other suppliers' service pages. The team now has a concrete access problem to resolve and a relevant question to measure again.&lt;/p&gt;

&lt;p&gt;That is a plausible contribution to the visibility gap. It is not proof that the access problem was the only reason for the omission. An external audit cannot read a model's internal decision process. Keeping that distinction explicit makes the diagnostic more useful and the conclusions easier to defend.&lt;/p&gt;

&lt;h2&gt;
  
  
  What makes this combination worth considering
&lt;/h2&gt;

&lt;p&gt;The practical advantage of B2A Radar is the connection between the evidence and the work that follows it:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Choose buying questions relevant to the business, with a defined market and language.&lt;/li&gt;
&lt;li&gt;Inspect mentions, recommendations, competitors and citations across models.&lt;/li&gt;
&lt;li&gt;Use the footprint report to identify technical and content problems worth fixing.&lt;/li&gt;
&lt;li&gt;Keep the completed reports and compare later checks after changes.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;For a small business or agency, having the full site check included in the monitoring subscription makes that workflow easier to use regularly.&lt;/p&gt;

&lt;p&gt;This is a product-fit argument, not an independent accuracy benchmark. Other platforms also offer analysis and recommended actions. A claim that one tool is more accurate than every competitor would require a shared test set, independently labelled answers and comparable measurements. We are not presenting such a benchmark here.&lt;/p&gt;

&lt;h2&gt;
  
  
  The free check and the subscriber report
&lt;/h2&gt;

&lt;p&gt;The public free check is a &lt;strong&gt;partial assessment&lt;/strong&gt;. It gives a starting view of the problems found in that check.&lt;/p&gt;

&lt;p&gt;For subscribers, the &lt;strong&gt;full site-check report is included at no additional charge&lt;/strong&gt;. It contains all issues detected within the scope of the completed check, rather than the partial public assessment. The report is saved in the account, with export and comparison options.&lt;/p&gt;

&lt;p&gt;The monthly allowances currently shown on the product page are:&lt;/p&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;Plan&lt;/th&gt;
&lt;th&gt;Included site checks per month&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Basic&lt;/td&gt;
&lt;td&gt;2&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Starter&lt;/td&gt;
&lt;td&gt;4&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Pro&lt;/td&gt;
&lt;td&gt;8&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;Full does not mean unlimited checks or a guarantee that an automated audit can detect every possible defect on any website. It describes the complete findings from that assessment. Monitoring questions and checking the site are complementary activities.&lt;/p&gt;

&lt;h2&gt;
  
  
  See the workflow
&lt;/h2&gt;

&lt;p&gt;This 72-second introduction shows the product:&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/Ik19BvMt4Uc" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;p&gt;Explore &lt;a href="https://b2aradar.com/" rel="noopener noreferrer"&gt;B2A Radar&lt;/a&gt;, the &lt;a href="https://b2aradar.com/en/brand-check" rel="noopener noreferrer"&gt;public partial check&lt;/a&gt;, and the &lt;a href="https://b2aradar.com/demo" rel="noopener noreferrer"&gt;sample report&lt;/a&gt;. The sample report uses demonstration data and can be viewed without signing up.&lt;/p&gt;

&lt;p&gt;The result to aim for is a repeatable process: identify a visibility gap, inspect the evidence, fix a confirmed issue, and measure again with a comparable set of questions.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>seo</category>
      <category>saas</category>
      <category>showdev</category>
    </item>
    <item>
      <title>AI visibility: mentions, recommendations and citations need separate checks</title>
      <dc:creator>Brand Design Ltd.</dc:creator>
      <pubDate>Wed, 18 Nov 2026 10:52:10 +0000</pubDate>
      <link>https://dev.to/branddesignltd/ai-visibility-mentions-recommendations-and-citations-need-separate-checks-lo1</link>
      <guid>https://dev.to/branddesignltd/ai-visibility-mentions-recommendations-and-citations-need-separate-checks-lo1</guid>
      <description>&lt;p&gt;When a buyer asks an AI assistant which provider to choose, a brand can appear in several very different ways. Treating every appearance as a recommendation makes a visibility report harder to act on.&lt;/p&gt;

&lt;p&gt;We built B2A Radar at Brand Design Ltd. to help businesses inspect those differences across GPT, Claude, Gemini and Perplexity. It is our own SaaS product. Here is the practical distinction behind the questions we ask.&lt;/p&gt;

&lt;h2&gt;
  
  
  Three signals to inspect
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;A mention:&lt;/strong&gt; the answer names your brand. It may simply list it alongside competitors, or discuss it without endorsing it.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;A recommendation:&lt;/strong&gt; the answer explicitly suggests your brand as an option for the buyer's stated need. The surrounding wording matters.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;A citation:&lt;/strong&gt; the answer links to a page as a source. That page may be yours, a competitor's or a third party's. Being named and being cited are separate observations.&lt;/p&gt;

&lt;p&gt;A useful report lets you inspect the answer and the source page, rather than leaving you with a score alone.&lt;/p&gt;

&lt;h2&gt;
  
  
  Keep comparisons meaningful
&lt;/h2&gt;

&lt;p&gt;Start with a defined set of buying questions. Record the country and language for each, and keep those settings consistent when comparing results over time. Separate the models so that a change in one assistant is not hidden by an aggregate.&lt;/p&gt;

&lt;p&gt;Read the competing brands and cited pages alongside your own results. They can help identify missing explanations, unclear positioning or information worth checking on your site. A change in a sampled answer is a reason to investigate, not proof that a single website edit caused it.&lt;/p&gt;

&lt;p&gt;Any visibility measure describes the questions, models and checks included in that measurement. It is not a universal ranking across every AI conversation.&lt;/p&gt;

&lt;h2&gt;
  
  
  What B2A Radar does
&lt;/h2&gt;

&lt;p&gt;B2A Radar brings mentions, recommendations, competitors and cited sources into one place. Scheduled checks show changes over time, and reports include suggested actions to review next.&lt;/p&gt;

&lt;p&gt;Here is a short introduction:&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/Ik19BvMt4Uc" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;p&gt;&lt;a href="https://b2aradar.com/?utm_source=devto&amp;amp;utm_medium=organic_social&amp;amp;utm_campaign=radar_launch_20261118" rel="noopener noreferrer"&gt;Explore B2A Radar and the sample report&lt;/a&gt;. Monthly monitoring plans start at €39. The sample report is available without sign-up and uses demonstration data.&lt;/p&gt;

&lt;p&gt;There is also a separate free check of whether AI services can read your website. Ongoing brand visibility monitoring is part of the paid subscription.&lt;/p&gt;

&lt;p&gt;How do you distinguish a passing brand mention from a useful recommendation in your own evaluations?&lt;/p&gt;

</description>
      <category>ai</category>
      <category>saas</category>
      <category>showdev</category>
      <category>analytics</category>
    </item>
    <item>
      <title>How We Keep an AI Agent's Payment Tied to an Accepted Offer</title>
      <dc:creator>Brand Design Ltd.</dc:creator>
      <pubDate>Wed, 30 Sep 2026 15:23:50 +0000</pubDate>
      <link>https://dev.to/branddesignltd/how-we-keep-an-ai-agents-payment-tied-to-an-accepted-offer-38fc</link>
      <guid>https://dev.to/branddesignltd/how-we-keep-an-ai-agents-payment-tied-to-an-accepted-offer-38fc</guid>
      <description>&lt;p&gt;A supplier page can describe a product. It cannot approve a payment.&lt;/p&gt;

&lt;p&gt;Suppose an agent is asked to buy a replacement server for €1,000. The supplier's page contains hidden text telling the agent to add a €500 consulting fee. If the agent submits a €1,500 checkout request, valid JSON is beside the point. The buyer never authorized the extra fee.&lt;/p&gt;

&lt;p&gt;This distinction shaped the business-to-agent (B2A) commerce flow we built at Brand Design Ltd., a branding and B2A agency in Varna, Bulgaria. An agent needs room to find and compare offers. The merchant needs a dependable record of the offer, its acceptance, and the buyer's separate permission to spend.&lt;/p&gt;

&lt;p&gt;On 22 September, six banks, including Bank of America and NatWest, published &lt;a href="https://newsroom.bankofamerica.com/content/newsroom/press-releases/2026/09/global-banks-collaborate-on-principles-for-trusted-agentic-comme.html" rel="noopener noreferrer"&gt;principles for trusted agentic commerce&lt;/a&gt;. The &lt;a href="https://www.w3.org/2026/ecommerce-agents/agenda.html" rel="noopener noreferrer"&gt;W3C and GS1 workshop&lt;/a&gt; earlier that month included sessions on agent identity, authorization and payment credentials. For a merchant, those concerns have to become checks at each purchase step.&lt;/p&gt;

&lt;h2&gt;
  
  
  Record the offer before accepting it
&lt;/h2&gt;

&lt;p&gt;Our direct purchase path follows five steps:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;catalog.read -&amp;gt; order.create -&amp;gt; order.accept -&amp;gt; payment.start -&amp;gt; order.read
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;ol&gt;
&lt;li&gt;
&lt;code&gt;catalog.read&lt;/code&gt; gives the agent the current service, price rules, currency and terms.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;order.create&lt;/code&gt; records a specific offer and an archived copy of the service page. A SHA-256 &lt;code&gt;pageHash&lt;/code&gt; identifies those archived bytes.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;order.accept&lt;/code&gt; checks the offer version and terms hash. If either changed, the agent needs a new offer and acceptance.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;payment.start&lt;/code&gt; checks the concrete payable amount and currency against the accepted offer and the buyer's authorized payment capability, including its limit and expiry.&lt;/li&gt;
&lt;li&gt;
&lt;code&gt;order.read&lt;/code&gt; returns the resulting order and payment state.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;The payment service makes these checks against recorded state. It does not use the model's explanation of a supplier page as evidence that the buyer approved a new price. The full path is described in our &lt;a href="https://github.com/NIKOMISHEV/b2a-agentic-commerce-spec" rel="noopener noreferrer"&gt;reference repository&lt;/a&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  Domain control is not spending authority
&lt;/h2&gt;

&lt;p&gt;We use a DNS TXT challenge to check control of a registered agent's declared domain. Current domain verification is required for autonomous shared-token payments by self-registered agents and for authenticated UCP access. It helps us reject an unauthenticated claim to represent that domain.&lt;/p&gt;

&lt;p&gt;That check does not identify the buyer or grant the agent a budget. The right to pay comes from a separate buyer-authorized capability. Even an authenticated agent cannot turn a line of untrusted page text into a higher spending limit.&lt;/p&gt;

&lt;h2&gt;
  
  
  What the hash example actually shows
&lt;/h2&gt;

&lt;p&gt;The repository includes a synthetic order manifest and a short Node.js verifier. It hashes a fixed set of offer fields and compares the result with a trusted hash supplied separately from the manifest. If an attacker changes both the amount and the hash inside the same untrusted JSON file, the separate commitment still fails.&lt;/p&gt;

&lt;p&gt;The sample &lt;code&gt;quote_hash&lt;/code&gt; is an illustration, not a production wire field or a digital signature. The published implementation uses &lt;code&gt;pageHash&lt;/code&gt;, offer versions and terms hashes. HTTP request authentication and signed route evidence are separate controls. None of these hashes, on its own, proves that the buyer consented to pay.&lt;/p&gt;

&lt;h2&gt;
  
  
  Where ACP, UCP and AP2 fit
&lt;/h2&gt;

&lt;p&gt;Our merchant system has deployed adapters for ACP and two dated versions of UCP. They use the same catalogue and order records as the direct route. Our UCP business profile is published on the merchant's own domain at &lt;a href="https://branddesign.ltd/.well-known/ucp" rel="noopener noreferrer"&gt;branddesign.ltd/.well-known/ucp&lt;/a&gt;. The route used for a transaction must be recorded and checked; merely having an adapter does not prove that a purchase ran through a platform's native checkout.&lt;/p&gt;

&lt;p&gt;Since 9 October, our UCP checkout also accepts AP2 checkout mandates. We sign the checkout, and before any charge we verify the mandate: the issuer's signature, the holder's signature, and whether the amounts and lines match the current checkout. We do not accept AP2 payment mandates or open mandates.&lt;/p&gt;

&lt;p&gt;There are platform limits too. &lt;a href="https://developers.openai.com/commerce/guides/key-concepts" rel="noopener noreferrer"&gt;OpenAI says&lt;/a&gt; building with ACP is open, while Instant Checkout in ChatGPT is currently available to approved partners. The &lt;a href="https://ucp.dev/2026-04-08/specification/checkout/" rel="noopener noreferrer"&gt;UCP checkout specification&lt;/a&gt; calls for a trusted user interface to finalize checkout unless the AP2 Mandates extension is supported. Google Pay is connected to our UCP checkout in test mode until Google approves the integration. We do not treat protocol support as permission to skip either platform onboarding or the buyer's authorization step.&lt;/p&gt;

&lt;h2&gt;
  
  
  Read-only access for agents
&lt;/h2&gt;

&lt;p&gt;Our MCP server at &lt;code&gt;branddesign.ltd/mcp&lt;/code&gt; is listed in the official &lt;a href="https://registry.modelcontextprotocol.io/v0/servers?search=branddesign" rel="noopener noreferrer"&gt;MCP Registry&lt;/a&gt; as &lt;code&gt;ltd.branddesign/commerce&lt;/code&gt;. In browsers that support WebMCP, every page of branddesign.ltd registers seven read-only tools for services, prices, the catalogue, evidence and capabilities, and a &lt;a href="https://branddesign.ltd/agent/CLI.md" rel="noopener noreferrer"&gt;command-line client&lt;/a&gt; gives the same reads. None of these read paths can create an order or start a payment.&lt;/p&gt;

&lt;h2&gt;
  
  
  What we have proved so far
&lt;/h2&gt;

&lt;p&gt;On 21 September, an external participant gave its agent a neutral task to buy a test service. The agent was not given our name, domain or checkout link. It found &lt;code&gt;101Ts3t&lt;/code&gt;, paid EUR 0.99 with its authorized wallet and completed the order. The &lt;a href="https://api.dothat.quest/api/verifications/qst_44e153fc16fe3fa91520e5f6922ae4c9" rel="noopener noreferrer"&gt;public verifier&lt;/a&gt; returned PASS for discovery, evidence integrity, transaction evidence and the completed agentic-commerce run. The proof package is in the &lt;a href="https://doi.org/10.5281/zenodo.22878834" rel="noopener noreferrer"&gt;version 1.3 Zenodo record&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;That is evidence for one real purchase. It is not a prompt-injection stress test, a reliability study or a live ACP/UCP checkout test. The &lt;a href="https://doi.org/10.5281/zenodo.23038829" rel="noopener noreferrer"&gt;version 1.4 record&lt;/a&gt; documents the added protocol routes, DNS checks, signed route evidence and synthetic regression checks. It does not report a new route-qualified live purchase. AP2 mandates, the UCP profile on branddesign.ltd and WebMCP on every page came after version 1.4.&lt;/p&gt;

&lt;p&gt;Offer integrity also does not prevent a duplicate charge after an ambiguous payment response. That failure mode needs idempotency and payment-status reconciliation. Our published hash example and the version 1.3 PASS do not establish those retry guarantees.&lt;/p&gt;

&lt;h2&gt;
  
  
  Verify it yourself
&lt;/h2&gt;

&lt;p&gt;Every item above has a public address you can open without an account. Our &lt;a href="https://branddesign.ltd/b2a-implementation-record" rel="noopener noreferrer"&gt;implementation record&lt;/a&gt; lists what runs today, from the direct API to UCP with AP2 mandates, MCP and WebMCP, with a link to check each one, the 21 September test and the DOI records.&lt;/p&gt;

&lt;p&gt;The agent can browse and choose. At &lt;code&gt;payment.start&lt;/code&gt;, the service still needs an accepted amount, a currency and the buyer's authority for that payment. If those do not match, the charge should not proceed.&lt;/p&gt;

&lt;p&gt;&lt;em&gt;Disclosure: This article was prepared with AI assistance and checked against the linked source records.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>architecture</category>
      <category>security</category>
      <category>fintech</category>
      <category>ai</category>
    </item>
  </channel>
</rss>
