This Week in Brand Leadership – Week 39 2026

Join the community

AnalyticsDev.net 2027 Conference

Register for updates and special offers

17 March 2027

Copenhagen

More info & Booking

Publication date: 22 September 2026
Reporting window: 15–21 September 2026

Executive overview

Who controls the customer relationship when software begins making more of the decisions? This week’s developments bring that question into focus.

Amazon’s dispute with Meta over AI shopping access and Pinterest’s new visual search advertising illustrate competing approaches to customer discovery. Salesforce’s CRM reasoning model brings the same issue inside the organisation: which actions should technology take, and who remains accountable? Meanwhile, scrutiny of shopping assistants’ product information highlights the importance of trust. Research on creative effectiveness and a new Indian marketing playbook provide a broader perspective on how brands should respond. Our editorial reading is that leaders need to connect automation with reliable information, distinctive ideas and clear decision rights. The priority is to establish where automation improves customer outcomes—and how teams will know when it does not. Sources and qualifications follow in each story.

1. Amazon’s clash with Meta exposes the struggle over AI shopping access

What happened
Reporting on 21 September said Amazon was blocking Meta’s Muse assistant from shopping on its marketplace and had asked Meta to remove Amazon access. Amazon cited unauthorised access and concerns about how the agent interacts with its service; Meta had not responded publicly in the cited reporting. Business Insider

Why it matters
For brands, an assistant’s ability to recommend a product does not guarantee that it can complete the purchase. Distribution planning will need to account for access agreements and retailer restrictions alongside discoverability.

BLC perspective
Treat agent-mediated shopping as a channel dependency to test. Ask which assistants can reach your products, which can transact, and what customer information returns to your organisation. A BLC discussion joining retailers, brands and technology providers could clarify where their incentives align—and where they conflict.

What to watch
Watch for negotiated access arrangements that determine which agents can transact and under what commercial conditions.

Sources

2. Pinterest makes visual discovery a more explicit performance channel

What happened
On 17 September, Pinterest announced Visual Search Ads, combining keyword relevance with prominent placements in search results and Pin closeups. The company said a beta would reach eligible advertisers in its advertising markets in the coming weeks; it also announced self-service A/B testing and additional product prioritisation controls. Pinterest Newsroom

Why it matters
This creates another potential investment point between inspiration and purchase. Creative assets and product information should be evaluated together: the image must attract interest, while the destination must fulfil the expectation it creates.

BLC perspective
The useful test is whether visual discovery brings additional customers or simply claims credit for existing demand. Set a clear acquisition or incremental-sales hypothesis before reallocating budget. Platform A/B testing can support learning, but its availability alone does not establish business incrementality.

What to watch
Look for independent advertiser results, especially new-customer acquisition and incremental profit rather than platform-reported return alone.

Sources

3. Salesforce brings specialised reasoning into CRM workflows

What happened
On 15 September, Salesforce and NVIDIA announced Koa, a CRM reasoning model built on NVIDIA Nemotron and trained with synthetic enterprise scenarios. Salesforce said it was entering selected customer pilots, with general availability expected in US regions in winter 2026; its claimed performance improvements come from its own benchmark. Salesforce announcement

Why it matters
The procurement question extends beyond whether AI produces a useful response. Leaders must assess whether it can perform the correct sequence of actions, respect permissions and recover safely when a workflow fails. Pilot design should include customer outcomes, exception handling and the cost of supervision.

BLC perspective
This connects directly with Jomar Reyes’s earlier MeasureCamp Czechia reflection: practitioners raised concerns about over-reliance on AI and the erosion of expertise needed to challenge its output. That discussion is community context, not evidence about Koa’s performance. As CRM work becomes more automated, someone still needs to understand why an action was taken. BLC — The Future of Brand Intelligence

What to watch
Watch for customer evidence on completed workflows, correction rates and operating cost, alongside availability outside the initial US rollout.

Sources

4. Shopping assistants face scrutiny over the product facts they surface

What happened
On 17 September, US senators Tammy Baldwin and Rick Scott asked the FTC to investigate Amazon and Walmart over allegations concerning country-of-origin information and misleading labels in AI-assisted shopping. This was a request for investigation, not a regulatory finding; Amazon denied intentionally withholding origin information and said it was improving its assistant, while Walmart declined to comment in the cited coverage. Senators’ announcement, The Wall Street Journal

Why it matters
An intermediary’s answer can shape perceptions of a brand’s provenance and credibility. Brands should test whether factual product attributes remain accurate when customers encounter them through an assistant.

BLC perspective
Product information management belongs in the brand-trust conversation. Establish who corrects inaccurate answers and what evidence supports the correction. A chatbot’s explanation of its own behaviour should not be treated as proof of a company’s intention.

What to watch
Watch for an FTC response and changes to how retailers validate and communicate product attributes through assistants.

Sources

5. WARC’s creativity research strengthens the case for sustained ideas

What happened
Coverage published on 18 September of WARC’s The Health of Creativity 2026 reported that 45% of ideas in the WARC Creative 100 also received effectiveness recognition, compared with 21% across creatively awarded ideas overall. The analysis draws on 6,360 ideas across 11 years of awards data and identifies an association between creative commitment and effectiveness. AdTech Today

Why it matters
The findings support a more demanding discussion about creative investment than asset volume or production speed. However, an awards dataset is selected evidence: it does not establish that winning a creative award causes commercial success or that these rates apply to all advertising.

BLC perspective
As content becomes cheaper to produce, leaders should protect the time and resources needed to develop a distinctive idea. For agency relationships, a useful question is whether the commercial model rewards enduring brand contribution alongside efficient delivery.

What to watch
Watch whether brands preserve campaign continuity while using AI to adapt execution, with commercial outcomes assessed over an appropriate period.

Sources

6. India’s DCODE 2.0 puts AI across the marketing organisation

What happened
DS Group, WPP Media and MICA launched the DCODE 2.0 marketing playbook on 17 September, integrating AI across search, paid media, content and commerce. Reporting on the launch highlighted practitioners’ concerns about fragmented measurement and the need to preserve emotional brand building while increasing execution speed. The Financial Express

Why it matters
The initiative offers an organisational signal from outside the US platform market: AI adoption is being considered across connected marketing disciplines. Its launch is evidence of an approach, not proof that the approach improves results.

BLC perspective
The brand–agency–academic collaboration is a useful model for practical education. BLC could apply the same principle to workshops built around one customer journey, bringing together creative, analytics and commercial teams to identify dependencies and test outcomes. Avoid transferring assumptions about Indian consumers directly to Nordic markets.

What to watch
Look for case studies showing how the framework changes decisions and improves customer or commercial outcomes.

Sources

BLC Radar

  1. European in-store retail media measurement is under review.
    IAB Europe’s September draft of its version 2 standards starts media measurement from footfall, clarifies exposure assumptions and proposes 28-day pre- and post-campaign sales reporting windows. Comments are requested by 23 October. This is an emerging standards item; the source does not establish a publication day within this edition’s reporting window. For brand buyers, the practical issue is comparability: ask what is observed, what is estimated and which assumptions drive the reported audience. IAB Europe — Draft in-store definitions and measurement standards
  1. Nordic marketing programmes are focusing on agent-mediated choice.
    RIMC’s programme for 17 September in Reykjavík centred on agentic commerce, with announced contributions from IKEA, Yoast and Nordic practitioners. The organiser’s advance release establishes the agenda, not conclusions or results from the event. This could provide useful speakers and practical cases for BLC discussions about product discoverability and customer choice. RIMC — Conference announcement
  1. AI education is expanding beyond prompting.
    The ANA’s programme for its 17 September Global Day of Learning included agents, synthetic audiences and cross-media measurement, with contributors including LEGO and Mastercard. Its announcement was published in August; the scheduled event falls within this week. The signal for BLC is an opportunity to connect practical AI skills with measurement literacy and professional judgement, then assess what participants can do afterwards. ANA — Global Day of Learning announcement

Question for the Community

As AI takes on more customer and marketing decisions, which decision would you insist remains accountable to a named person—and how would you know that person can still challenge the machine?

Homepage Overview

HOMEPAGE HEADLINE
Who Controls the Customer Relationship as AI Advances?

HOMEPAGE OVERVIEW
Amazon’s clash with Meta over shopping access, Pinterest’s visual search ads and Salesforce’s CRM reasoning model sharpen a leadership question: who controls customer decisions as AI becomes more active? This Week in Brand Leadership examines the implications for brand trust, marketing investment and organisational capability, alongside evidence on the continuing value of creative commitment.

About AI and this briefing

This Week in Brand Leadership is produced with AI assistance using a consistent BLC editorial framework. Alongside current research, the process draws on historical insights from Brand Leadership Community interviews, events, articles and practitioner discussions. AI supports research, analysis and synthesis, while the editorial perspective is shaped by BLC’s accumulated community knowledge and judgement.

Editorial disclaimer

The “BLC perspective” sections provide general editorial commentary for information and discussion only. They do not constitute professional advice tailored to your circumstances. Readers are responsible for independently assessing the information and obtaining appropriate advice before acting. To the fullest extent permitted by applicable law, Brand Leadership Community (BLC) accepts no liability for any loss or damage arising from reliance on these perspectives or actions taken on their basis.