Manuel Acevedo

Reporting on the Anthropic–Salesforce "Claudeforce" pilot, which they say was still in pilot with an open beta coming in September, they quote Salesforce's Patrick Stokes saying the company's value rises when people stop entering the CRM through a screen and start operating it through language.

• Their operational takeaway: if the CRM answers from inside with the same permissions as before, the question shifts from what the software can do to how a sales team measures its own work.
• They close by asking commercial managers in Peru and the region whether their team still counts hours spent in the CRM or decisions the CRM took on its own.
• Before relying on the capability, ask whether the beta has launched and what is available.

A sales manager who today opens six browser tabs to assemble a pipeline report can ask for it in one sentence and get the answer inside the CRM, under the same permissions.

Read original

Their reading: the change does not punish using AI on WhatsApp; it punishes using it badly.

• The key metric is turns per resolution: an agent that needs eight messages to resolve what a well-designed one resolves in three starts costing money on every extra exchange.
• Acevedo recommends auditing flows — fewer back-and-forth questions, direct answers, and a bot that knows when to hand the conversation to a person.
• The same change, they argue, improves the customer experience.

Acevedo compares the eight-message agent to paying a salesperson by the minute to circle a sale instead of closing it.

Read original

In their account, Stripe presented real agentic-commerce cases at the Mexico E-Commerce & Retail Summit 2026.

• Reporting from the event, they cite Stripe's claim that a purchase made through an AI agent converts four times better than a conventional e-commerce buyer because the agent arrives with the decision already made.
• Before using the figure, ask Stripe for its methodology.
• The consequence Acevedo draws is not a marketing tactic but a precondition: a brand has to be legible enough for the agent to choose it.
• Latin American companies still have time to prepare, they say, though the window is shorter than it looks.

An agent that compares prices, reads reviews, checks availability, and buys on the buyer's behalf — with product data and signals clear enough for it to pick one brand over another.

Read original

They read this as evidence that AI pays attention to what a brand's customers say.

• The LATAM-specific point is location: in Peru and the region, those reviews frequently sit on Mercado Libre, Falabella, or Google Maps rather than on the brand's own website.
• Before asking customers for more reviews, Acevedo says, a brand should find out where AI is actually reading them — a check they say can be done in an afternoon and offer to walk a team through.
• Ask for the study link to verify the figures.

Audit the marketplace, retail, and map profiles where a brand's reviews already live, rather than collecting new reviews first.

Read original

In the Claudeforce post, they say Collective Intelligence helps commercial teams in LATAM distinguish what can already be used from what is still a promise, and offer a reading "without smoke." The other posts follow the same template: a platform announcement or policy change, a plain-language explanation, a concrete operational consequence, and an offer to work through it.

Vendor announcements — the Claudeforce pilot, the Meta messaging change, agentic commerce from Stripe's session, and the AI-citation study — each become one thing a commercial, e-commerce, or service team should check or change.

Read original · Original 2 · Original 3 · Original 4

Start here to see how Manuel Acevedo reads a vendor announcement and turns it into a question for a sales team.

Manuel Acevedo reports that Anthropic and Salesforce announced Claudeforce, a pilot integration due to reach open beta in September.

• They quote Salesforce's Patrick Stokes saying the company's value rises when people stop entering the CRM through a screen and operate it through language instead.
• The example is concrete: a sales manager who today opens six browser tabs to assemble a pipeline report can ask in one sentence and get the answer from inside the CRM, under the same permissions as before.
• The post closes with a question for commercial managers in Peru and the wider region: does your team still measure productivity by hours spent in the CRM, or by decisions the CRM took on its own?

For brands planning a campaign

For a B2B SaaS vendor selling to LATAM commercial teams, this post shows where Manuel Acevedo draws the adoption line: not at the feature list, but at whether a team's metrics have caught up with automation. A pitch built around replacing screen work with conversational work fits this creator's market view better than a generic AI-productivity claim.

It shows why the creator treats machine readability as a commercial problem rather than a technical one.

Manuel Acevedo describes a buyer's AI agent comparing prices, checking reviews, verifying availability and, in many cases, completing the purchase without the buyer opening a browser.

• The post says Stripe presented real agentic-commerce cases at the Mexico E-Commerce & Retail Summit 2026 and that a purchase made through an AI agent converts four times better than a conventional e-commerce buyer because the agent arrives with the decision already made.
• For a brand, the consequence is legibility: it has to be readable enough for the agent to choose it.
• Acevedo adds that Latin American companies still have time to prepare, but the window is shorter than it looks.

For brands planning a campaign

A B2B SaaS pitch aimed at LATAM commerce and sales operations can use this post as a market-context anchor: the creator's position is that agent-mediated buying rewards product-data quality and machine readability, a different budget line from paid acquisition. The 4X figure and the Stripe session are reported from the event, not from the creator's own testing.

Manuel Acevedo says Meta will charge for service messages a business sends inside the 24-hour window on the WhatsApp Business API from 1 October.

• The change covers CRM, chatbot and automation setups once they exceed 1,000 free messages a month.
• The argument: this does not penalise using AI on WhatsApp; it penalises using it badly.
• An agent that needs eight messages to resolve what a well-designed one resolves in three now costs money on every extra turn.
• The post compares this to paying a salesperson by the minute to circle a sale instead of closing it.
• The recommended action is to audit flows: cut back-and-forth questions, give direct answers, and make sure the bot knows when to hand the conversation to a person.
• The same changes can improve the customer experience.

For brands planning a campaign

For a SaaS product that shortens or resolves customer conversations, this post makes the cost case: cost per resolved conversation becomes a defensible metric, and the creator has framed a leaner flow as both cheaper and better for the customer. A campaign asset quantifying turns saved per resolution would speak directly to the argument here.

It is the most directly actionable piece in this set and a good entry point for someone looking for a task rather than a trend.

Manuel Acevedo cites a study of 237,000 AI citations that found 59% of cited product listings had more than 100 reviews and 92% had four stars or more.

• They read the results as evidence that AI pays attention to what a brand's customers say.
• For Peru and LATAM, reviews often sit on Mercado Libre, Falabella or Google Maps rather than the brand's own site, so a brand should first find out where AI is reading reviews before asking for more.
• The post offers to run that audit with a team that writes in and says the check can be done in an afternoon.

For brands planning a campaign

This is the strongest angle for a B2B SaaS or martech offer built around AI-search visibility: the creator offers a stated, checkable starting point — find where AI reads your reviews before adding more — instead of a generic “optimise for AI” claim. A sponsored study or audit template that the creator could use to guide a LATAM brand fits the format already in use.

The post cites Salesforce executive Patrick Stokes saying that if people stop entering the CRM through screens and start operating it through language, Salesforce's value rises. Acevedo illustrates the shift with a sales manager who today opens six tabs to build a pipeline report and could instead ask Claude in one sentence, with the same permissions as before.

Uses a named executive quote, a concrete sales-manager scenario, and a diagnostic question for LATAM sales leaders.

The pilot and beta status limit the post's value as evidence of deployed capability.

For brands planning a campaign

For a B2B SaaS brief aimed at LATAM commercial teams, this post shows Acevedo translating an enterprise platform shift into an operational question about CRM productivity and decision-making.

• A campaign could use that angle to discuss how sales teams should measure work as AI-assisted workflows spread.
• The post does not prove Claudeforce is generally available: Acevedo describes it as a pilot, with open beta coming in September.

Examples: Source 1

The argument is that this does not punish using AI on WhatsApp but punishes using it badly. For example, an agent that needs eight messages to solve what a well-designed one resolves in three now creates a cost for every extra turn.

Uses a cost analogy about paying a salesperson by the minute and asks readers to audit their WhatsApp flows.

For brands planning a campaign

For a customer-experience or automation brief targeting LATAM sales and service teams, this post can support a campaign about auditing conversational flows, reducing back-and-forth, and designing clear handoffs to a person. It gives a cost-based reason to improve the flow rather than just add a bot.

Examples: Source 1

The post references Stripe cases presented at the Mexico E-commerce & Retail Summit 2026 and states that purchases through an AI agent convert four times higher than traditional e-commerce buyers because the agent arrives with the decision already made. The practical implication is that a brand must be legible enough for the agent to choose it.

Starts with a buyer-agent scenario, adds a conference case reference and a conversion figure, then turns to brand readiness.

Treat the conversion figure as Acevedo's cited claim, not a fully documented study.

For brands planning a campaign

For a retail or e-commerce brand in LATAM, Acevedo is relevant to campaigns about product data, availability, review signals, and agent readiness.

• The post frames preparation as urgent but still possible for Latin American companies.
• It states that purchases through an AI agent convert four times higher than traditional e-commerce buyers, without linking the methodology; present this as Acevedo's directional claim, not a verified benchmark.

Examples: Source 1

The plain-language conclusion is that AI pays attention to what customers say. Many brands in Peru and LATAM keep reviews on Mercado Libre, Falabella, or Google Maps rather than on their own websites; Acevedo suggests auditing where AI is actually reading those reviews before asking for more.

Offers a simple, time-boxed audit idea and an invitation to review it with a team.

Ask for the study and its method before using the figures as verified data.

For brands planning a campaign

For a reputation-management, AEO, or GEO brief in LATAM, this post shows Acevedo connecting AI search visibility to concrete review signals and platform locations.

• A campaign could use the audit idea to help brands review marketplace and map profiles, not only their own site.
• The post does not link the underlying study, so present the numbers as Acevedo's cited evidence rather than independently verified data.

Examples: Source 1

Recurring arguments include the shift from screens to language and agents, the need to improve AI-driven workflows rather than add more steps, and the importance of being readable to AI systems that compare and recommend. Collective Intelligence is presented as helping teams separate what is usable now from what is still a promise.

Uses a platform-news hook, a concrete example or figure, a diagnostic question, and an offer to talk.

The posts have no publication dates; ask for dated examples to assess how the topics or arguments have changed over time.

For brands planning a campaign

A brand can evaluate Acevedo as a practical, region-specific translator who connects vendor news to sales, e-commerce, and service operations. The posts are short LinkedIn commentary rather than long-form research, so a campaign should not present them as deep technical analysis.

Examples: Source 1 · Source 2 · Source 3 · Source 4

They use “we” to describe the firm's work but do not specify Acevedo's role. Ask what their relationship to the firm is.

Original disclosure

En Collective Intelligence ayudamos a equipos comerciales en LATAM a distinguir qué de esto ya se puede usar y qué todavía es promesa. Si quieres una lectura sin humo, hablemos.

View example

Both treat process and response speed as practical issues behind sales technology. Dalle Molle's posts on the five-minute lead-response window and on leads dying from poor follow-up make the same argument in a more operational register than Acevedo's discussions of CRM work and WhatsApp automation.

Dalle Molle's posts draw on direct operating experience and sit alongside Escala, a CRM and sales system.

• They focus on pipeline hygiene, automated follow-up, and the tool that carries it out.
• Acevedo more often starts with third-party platform announcements or studies, positioning Collective Intelligence as an advisory filter between vendor news and a team's next decision, rather than as the product doing the work.

Read original · Original 2

The Unbound post echoes Acevedo's concern about CRM access and automation: bad context produces worse results than no AI at all, and governance matters because agents read the data they are given.

Moena writes from a revenue-operations consulting and HubSpot-partner perspective, covering hiring, team design, paid-media roles, and event notes from an ecosystem partner day. Acevedo takes a wider view of vendor products and buyer behavior, from Salesforce and Anthropic to Stripe's agentic-commerce cases and Meta's WhatsApp pricing, framing the reader as a commercial or e-commerce operator deciding what to do next.

Read original · Original 2

Palacio's account of turning an online comics store into an AI-driven commerce experience addresses the same shift Acevedo describes from browsing to conversation. Both writers argue that AI becomes interesting when it solves a real customer or business problem instead of only answering questions.

Palacio builds and operates AI products, describing his own platforms across music, faith, guitar learning, and comics, and reports milestones such as awards and user numbers.

• Acevedo generally does not present themself as the builder of the systems they discuss.
• They interpret announcements, studies, and policy changes from Salesforce, Stripe, Meta, or third-party research, and ask what commercial teams should do about them.

Read original · Original 2

The posts argue from inside the sales process: a lead not contacted within five minutes is already cold, and the usual fault is a CRM that works as a contact list instead of a system that triggers the next action.

• One post ties slow response to a specific cost — the 78% of buyers who purchase from whichever company answers first — while the other closes by asking how many leads are disappearing unnoticed.
• A vendor selling lead routing, response automation, or WhatsApp follow-up gets a creator whose examples cover the exact mechanics the product changes.
• The same campaign decision — which creator carries a LATAM commercial-teams campaign for a B2B SaaS vendor — tips toward Andrea when the product is the CRM itself rather than an AI platform shift.

The range is narrower.

• Both posts make the same speed-to-lead and follow-up argument, giving the campaign one repeated proof point rather than a spread of topics.
• The work stays tactical, with no comparable commentary on platform moves at Salesforce, Stripe, or Meta for a broader SaaS story to borrow.
• A vendor seeking event coverage, market analysis alongside the operational case, or a wider topical surface gets less from Andrea than from Manuel.
• Before booking, ask for audience data and sponsorship terms.

Examples: Source 1 · Source 2

The reviewed work presents a RevOps consultancy and HubSpot Platinum Partner that describes itself as an architect of revenue.

• One post reports from HubSpot's Partner Day at UNBOUND26, quoting Yamini Rangan on the point that bad context is worse than not using AI.
• It then shows how bad CRM data can lead an agent to offer a discontinued product or route an approval to someone who changed roles, before adding two more operating ideas: buyers now test before committing, and most companies build agents without governing them.
• That speaks directly to the buyer's data-quality and process problems, while giving the campaign partner-ecosystem and event context the other options do not.

One post is a hiring announcement for two roles in Santiago, useful for showing how the team works but not for buyer education.

• The posts also take different forms — recruitment and event commentary — making the output less consistent than a single-thread argument.
• The partner-event angle depends on the vendor wanting HubSpot ecosystem context rather than the plain Spanish market commentary Manuel produces.

Examples: Source 1 · Source 2

The posts describe creating MixingMusic.ai, CatolicosGPT, and GuitarraIA, and rebuilding an online comics store as an AI conversation commerce experience.

• They name the Anthropic and OpenAI APIs and report adoption figures and awards.
• The commerce post argues the shift directly — no navigating, conversing; no searching, discovering — and says the coming contest is won by whoever has more intelligence, not more products.
• A vendor selling AI product infrastructure or conversational commerce gets an operator's account of what it takes to launch, train, and ship these products: a different, more concrete credential than commentary about other companies' platforms.

The content focuses on project showcases and promotion, rather than the platform-change-to-operating-decision arc a sales-team reader follows.

• Its topics — music production, faith, guitar learning, comics — sit further from CRM, pipeline, and revenue operations.
• For a LATAM commercial-teams campaign, the angle would need to shift to AI product building rather than the sales process Manuel writes about, putting the creator on a different problem from the one the vendor solves.

Examples: Source 1 · Source 2