Matti Perkkiö

ChatGPT and Claude learn about a person as they are used, while ERP and CRM systems already contain customer and company data.

• But those systems do not know which facts matter most to a given user.
• Moving from agents that handle narrow tasks to software that advises a person requires more context.
• The bottleneck, Perkkiö says, is not model quality but how much useful intelligence has been provided, and how well it covers the language and domain in which the work happens.

Perkkiö points to the company's phone agents, which answer customers on a user's behalf. To do that convincingly, they need enough material to handle customer language, customer problems and construction-industry vocabulary in Finnish, Swedish and Norwegian.

Read original

Good tools make companies document their sales playbook and define how their products genuinely differ from competitors, support every phone or remote meeting and its preparation, analyse what happened afterwards, and remove paperwork from a salesperson's week.

Reacting to Faction's "Myynnin tila 2026" report, Perkkiö moves beyond generalities and lists concrete steps: use Revial.ai to establish a sound sales process and support meetings and their preparation; if it is not in use, start with any transcription tool that produces meeting notes, a to-do list and CRM entries; and use Easoft's Tarjousagentti to generate a quote and tailored customer landing page without hours of paperwork. The post carries a 20 percent discount code for a trial and invites readers to message Perkkiö about AI quote agents and other integrated agents.

Perkkiö tells readers to forget ChatGPT, yet elsewhere treats ChatGPT and Claude as examples of systems that learn from use. The apparent objection is to tools and manual work detached from the sales process, not to general assistants as such.

Read original · Original 2

AI is the right answer for routine work, not judgement. The episode is pitched at leaders who want to use AI for growth and customer success.

Perkkiö says the conversation draws on 25 years in the construction industry and appears alongside Ashwin Sivakumar and Janne Vainikainen. Ask for the episode to assess the other participants' arguments; only Perkkiö's announcement text is available here.

Read original

Turning software into an assistant that advises you depends on that input. The practical task is persuading staff to use the systems the company already owns, not just buying better tools.

A post opens with a light observation about LinkedIn's own engagement games, then asks how to get people to enter valuable data into a CRM or ERP with the same enthusiasm. Perkkiö also promotes a Finnbuild coffee session on AI in construction and renovation operations management, capped at the first fifty sign-ups, as a live conversation for operators.

Read original

Open with this to see the problem being solved, not just an opinion about AI.

The post opens lightheartedly about LinkedIn's own engagement games, then turns to a serious question: how do you get staff to put valuable data into a CRM or ERP the way they happily engage with a game?

• Perkkiö argues that AI's biggest limitation is not the model but how much useful intelligence it has been given, describing this as an "intelligence layer" or "second brain."
• CRM and ERP systems already contain customer and company data, but they do not know which facts matter most to a given user.
• Getting from agents that do narrow tasks to software that advises you, they argue, requires feeding it more context.
• The concrete example is the company's phone agents, which answer customers on a user's behalf and therefore must understand customer language, customer problems, and construction-industry terminology in Finnish, Swedish and Norwegian.
• The post closes with an invitation to a Finnbuild coffee session on AI in construction and renovation operations management, capped at fifty sign-ups.

For brands planning a campaign

Most useful as a webinar or podcast spine: the argument that AI adoption stalls on data quality and domain vocabulary, illustrated with multilingual construction phone agents, is a ready-made talk for a B2B software audience. The Finnbuild invite shows Perkkiö already uses in-person events as a channel, making event tie-ins a realistic format alongside studio content.

Before judging an AI-in-construction claim, a reader wants to know whether the person has worked in the industry. Perkkiö states that background directly and explains what they think AI is and is not for.

Perkkiö announces a guest appearance on the Signal to Scale podcast alongside Ashwin Sivakumar and Janne Vainikainen, framed around how AI changes marketing and scaling.

• They say the discussion draws on 25 years in the construction industry and stress two positions: understanding customer needs matters more than the tools, and human connections and strategic differentiation drive growth.
• AI is presented as the right answer for routine work rather than judgement.
• Ask for the episode link to review the discussion before booking.

For brands planning a campaign

Supports a guest-expert format rather than a creator-style sponsorship.

• The hook is a construction veteran telling sales and operations leaders which parts of their work AI should take over and which it should not.
• That fits a B2B software firm's webinar or podcast series for construction and sales leaders.

Readers new to Perkkiö can see how they turn a research report into recommendations a sales team can act on.

Reacting to Faction's "State of Sales 2026" report, Perkkiö turns the findings into specific actions for a sales organisation.

• First, adopt Revial.ai to build a sound sales process from the start: the tool forces a company to write a sales playbook and define how its products genuinely differ from competitors, then helps with every phone or remote meeting and preparation for one.
• Afterwards, it analyses what could be improved and what the customer actually said.
• Second, if Revial is not in use, start with any transcription tool that produces meeting notes and a to-do list, then pushes them into the CRM.
• Third, reduce a salesperson's paperwork with Easoft's Tarjousagentti, which generates a quote and a tailored landing page for the customer.
• Perkkiö recommends dropping ChatGPT and spreadsheets in favour of sales AI tools that integrate directly into the sales and sales-management process.
• The post also offers a 20 percent discount code for Revial's free trial and invites readers to message them about AI quote agents and other agents integrated with Easoft.

For brands planning a campaign

Shows an established pattern of walking a sales audience through named tools, including a discount code and an open invitation to talk—a useful model for a B2B software vendor considering co-marketing. The post's structure, a report finding followed by a short ranked list of actions, gives a sponsor a way to feature a product without making the post read like an ad.

One post walks through Faction’s “Myynnin tila 2026” report on how respondents use AI in sales, then offers concrete steps: use a tool that makes a company document its sales playbook and define product benefits against competitors, then supports every phone or remote meeting and its preparation; use a transcription tool to produce meeting notes, to-do lists and CRM entries; and use a quoting agent to draft an offer and a tailored landing page for the customer.

• Perkkiö’s closing advice is to drop ChatGPT and spreadsheets in favour of sales AI tools that integrate directly into sales and sales management, which they say is where competitiveness and control come from.
• They also offer a discount code for a trial and invite direct messages about integrated agents.

Perkkiö lays out the case in a short numbered list of practical tools and steps, then adds an emphatic bonus recommendation. Readers leave with actions, not just a survey of the market.

For brands planning a campaign

A vendor selling sales software, quoting tools or AI agents could ask Perkkiö to show where a tool fits in the sales process and what it takes off a rep’s plate, rather than deliver a general talk on AI trends. The discount-and-message call to action also shows they are comfortable promoting a trial offer in a collaboration.

Examples: Source 1

They call this an “intelligence layer” or “second brain”: ChatGPT and Claude learn about users as they use them, and ERP and CRM systems already hold customer and operations data, but those systems do not know what matters most to a given person.

• Moving from using agents to having software act as an assistant that offers advice requires feeding in more information.
• Perkkiö illustrates the point with phone agents that answer customers on a user’s behalf and need enough content to handle customer language, customer problems and construction-industry vocabulary in Finnish, Swedish and Norwegian.
• The post also promotes a Finnbuild coffee session on AI in construction and renovation operations management, capped at the first 50 sign-ups.

Perkkiö moves from a light observation about LinkedIn engagement to a broader claim about how AI systems learn, then grounds the point in their own agent product and a named industry event.

For brands planning a campaign

This is a strong hook for a webinar or podcast segment aimed at construction and operations leaders: how to get value from data already in CRM and ERP systems, and what it takes to make voice or chat agents work in a specialised industry and in more than one language.

Examples: Source 1

They say they share insights from 25 years in the construction industry, emphasise the power of human connections and strategic differentiation, and frame the episode around why understanding customer needs and using AI for routine tasks matter for growth. The post targets leaders looking to use AI for growth and customer success.

The promotional post does not include the episode itself. Ask for the episode before booking if you need to assess the conversation’s specific arguments.

For brands planning a campaign

Useful for a software company seeking a voice credible with construction and renovation firms, not just AI specialists. A co-hosted conversation about adoption, differentiation and the routine work AI can absorb fits the position Perkkiö presents here.

Examples: Source 1

The mix points to a Finnish-speaking practitioner audience in construction, sales and business software, with a smaller English-language reach through the podcast.

For brands planning a campaign

A brand can choose the format to suit its needs: a Finnish-language post or event for a Finland-focused construction and sales audience, or an English-language, podcast-style conversation for a broader leadership audience. The event post shows Perkkiö can drive sign-ups for a live session with a stated capacity.

Examples: Source 1 · Source 2 · Source 3

They also share a discount code carrying Easoft's name.

Original disclosure

Ps. jos otat Revialin ilmaiseen testikäyttöön, saat koodilla EASOFT20 20% alennuksen. Jos AI-tarjousagentit ja muut myyntiin Easoftissa integroituvat agentit kiinnostavat, laita ihmeessä minulle viestiä, niin sparraillaan lisää.

View example

They offer a 20% discount code for its free trial.

Original disclosure

Ps. jos otat Revialin ilmaiseen testikäyttöön, saat koodilla EASOFT20 20% alennuksen.

View example

View example

They discussed customer needs, using AI for routine tasks, human connections in business, and strategic differentiation, drawing on 25 years in the construction industry.

View example

Lukeš describes Vodafone Verify checking identity and AI-generated content, and INER providing emergency connectivity.

• Perkkiö writes about AI agents that need domain knowledge and AI embedded in the sales process.
• Both see adoption as a matter of building systems people can rely on.

Lukeš’s examples are cybersecurity, digital identity and emergency response for telecom and public institutions, often in Czech. Perkkiö’s are construction, sales operations and CRM/ERP data for Finnish and Nordic business audiences.

Read original · Original 2

McLellan’s Build a Better Agency episodes examine agency partnerships and leadership follow-through.

• Perkkiö appears on Signal to Scale to discuss AI, customer needs and construction experience.
• Both turn experience into frameworks for leaders.

McLellan focuses on agency ownership, partnership choices and leadership credibility for marketing agency leaders. Perkkiö focuses on AI adoption in sales and construction operations, for a Finnish-language practitioner audience.

Read original · Original 2

Snow argues that AI exposes process problems and that visibility matters more than effort. Perkkiö argues that sales AI should sit inside the sales process and that systems need user knowledge to be useful.

Snow focuses on sales-team follow-up, CRM visibility and coaching in North American B2B sales. Perkkiö focuses on AI integration across sales and operations, with construction-industry examples and Nordic languages.

Read original · Original 2

Snow's two posts work the identical problem — AI landing on a sales process that does not drive action — but stay inside the sales organisation.

• One argues that telling reps to "be more organized" signals a broken process, and that AI does not fix execution problems, it exposes them: teams getting results use it to tighten the system, not to add activity.
• A companion post makes the CRM point concrete, contrasting a handwritten to-do list with tasks buried behind dashboards and reports, and tracing drifting deals to next steps that are easy to miss.
• For a sales-engagement or follow-up product aimed at sales leaders, Snow is the closer subject match.

The tradeoff is language, industry and format.

• Snow's posts are in English, written for a general sales audience, with no construction or renovation context and none of the multilingual customer vocabulary Perkkiö's phone-agent example turns on.
• A Finnish or Nordic construction-software vendor selling to builders reaches those buyers more directly through Perkkiö's Finnish industry framing.
• Nothing in Snow's two posts shows a live or recorded session format, so a sponsor would be buying a written argument and its audience rather than a webinar or podcast slot with questions, a walkthrough and a stated registration cap.

Examples: Source 1 · Source 2

The two items describe episodes of Build a Better Agency: a solo episode on whether an agency owner should take a business partner, offering nine non-negotiables plus self-reflection prompts and alternatives such as compensation or peer support; and an interview with Steve Farber on leadership credibility, unintentional promises, a weekly journaling exercise for tracking commitments and inviting team feedback on follow-through.

• Both show a settled format — a named guest, one defined problem, a framework the listener can use the same week — which is the shape a sponsor would step into.
• Choose McLellan if the sponsor's need is partnership, trust or agency operations, or if the sponsor needs a long-running show with an established host.

The tradeoff is subject and audience.

• Neither episode touches AI integration in sales or operations, and the content is aimed at agency owners rather than construction firms or sales organisations, so the practical question Perkkiö's posts answer — where in the process a tool plugs in and what data the CRM or ERP needs — has no counterpart here.
• Both items are episode announcements rather than episode transcripts, so ask for access to the recordings before judging how the host performs with a sponsor.
• Language also differs: this is English-language agency content against Perkkiö's Finnish industry content.

Examples: Source 1 · Source 2

Both posts describe Vodafone's own service work.

• The first sets out a family of verification services: an app for checking text, images, video and audio with AI analysis for manipulation and deepfakes, a tool for verifying participants in online meetings and detecting substituted people, and detection of synthetic signals during phone calls, introduced alongside named Vodafone Business executives.
• The second moves to resilience, describing the INER emergency-response kit — a portable set that restores satellite data, local Wi-Fi and connectivity where power is unavailable — presented with Justin Waller at the IS2 security summit.
• Choose Lukeš if the sponsor sells identity verification, security or enterprise communications and needs a Central European technical audience, where the content sits on a conference platform as well as in posts.

The tradeoff is that this is a different conversation from the one Perkkiö's posts have.

• The posts do not cover sales process, construction operations, or how a company feeds its existing CRM and ERP data into AI agents, so a sponsor selling into construction sales or sales execution would not reach the same buyers.
• The AI subject here is framed as risk — deepfakes, forged identity, false communication — rather than as a practical integration decision, and the readers are security and enterprise-technology people.
• The posts are written about a named employer's product line, which a sponsor should weigh before expecting an independent recommendation, and they are in Czech, against Perkkiö's Finnish material.

Examples: Source 1 · Source 2