Seven Companies, One Team: Leading People Through Post-Merger Integration with Data and AI

Issue 017 · Published 13 October 2026 · The AI Executive Briefing

IN ONE MINUTE

Most change programmes stall on people, not plans. Gartner's 2025 research found that only 45% of employees achieved the change goals their organisation had set. As General Manager, I held full P&L responsibility for a global business of about US$800 million in annual turnover, built from seven acquisitions that still behaved like seven companies. I did not start with systems. I started with respect for the people who had built those businesses, then used data, technology and a machine-learning (ML) pricing engine to give everyone the same facts and the same direction. Over the 12-month programme, sales grew 11%, finishing 6.7% above the planned sales figure; profit grew 16%; operating costs fell 14%; and operating efficiency improved 17%.

THE OBJECTIVE

Ask leaders what makes change hard and most will say people. They are right. But people rarely resist change because they are difficult. They resist when they cannot see where it leads, what it means for them, or whether they will be judged fairly.

Technology cannot replace leadership. What it can do is give people transparency and direction, and keep every decision tied to the business strategy. AI raises the stakes on both sides: introduced well, it makes decisions far more transparent; introduced without trust, it becomes one more thing people resist. This issue shows how that worked in one integration, at three levels of the organisation, and what it means for each member of the leadership team.

WHAT THE RESEARCH SAYS

Gartner's 2025 research shows how hard change has become. In a survey of more than 2,500 employees, only 45% said they had achieved the change goals set by their organisation. In a parallel survey of nearly 1,000 mid- to senior-level leaders, 36% said their teams tend to hesitate and delay action to see whether a change will stick, and 39% described change as a significant source of stress for their teams.

Acquisitions raise the stakes. In Bain & Company's M&A Practitioners' 2023 Outlook Survey, nearly half of respondents cited cultural fit or difficulty integrating management teams as a primary reason their past deals failed.

AI is now entering the most sensitive commercial decisions. McKinsey's November 2025 survey of 419 pricing leaders found that 65–85% of organisations expect to adopt generative or agentic AI in pricing within one to three years, up from 10–30% today.

The lesson is consistent. The plan is rarely the problem. Adoption is.

THE CASE: SEVEN COMPANIES, ONE TEAM

The situation

The business spanned luxury products, cosmetics, packaging and digital retail solutions. It served Fortune 500 clients and worked with more than 1,000 customers and suppliers. Its capabilities were strong. Its structure was not. Each acquired company kept its own P&L, customers, processes, suppliers, systems and reporting, and five former owners sat in the leadership team.

The symptoms were familiar to anyone who has led an integration. Sales teams pursued the same Fortune 500 accounts. Teams priced with different assumptions and margin definitions. Procurement bought similar materials separately. Finance spent its time reconciling numbers before it could report them. Leadership meetings became arbitration: whose customer, whose revenue, whose forecast.

Nobody was underperforming. Everyone was optimising for their own business, because there was no shared business to optimise for.

Starting with respect

The first thing I brought into the room was not a plan. It was respect. The former owners and business veterans knew far more than I did about their supply chains, operations and customers. They also had legitimate interests to protect. Like most leaders of acquired companies, they were working to deliver the performance agreed in the acquisition deal and to maximise their earn-outs.

Ignoring those interests would have guaranteed resistance. So I put them on the table. I was transparent about what the business needed, then worked with the holding company and other stakeholders on a plan that realigned priorities so the integration created value for every party, the former owners included. Those were not short or easy conversations. But they were the only route to a goal everyone could commit to.

The vision we agreed was simple:

To become one integrated, customer-centric and digitally enabled business partner for Fortune 500 clients, combining global supply-chain execution, innovation and transparent performance management.

Leading change at three levels

Resistance looked different at each level of the organisation, so the response had to be different too.

• Senior executives needed to understand and agree the changes. That meant respecting their experience and standpoints, debating openly, and resolving management decisions together rather than imposing them.
• Middle managers needed to see the value for themselves: how the new operating model would help them perform better, what was now expected of them, and full visibility of their KPIs, so that no one would fall behind without warning.
• Operating teams needed to know why the change was happening, followed by proper training and a realistic adoption plan, so they could move to new ways of working with confidence and trust.

Before redesigning anything, our integration team mapped the real workflow with every function, from customer request to cash collection. Some local practices reflected genuine expertise and were kept. Others were workarounds for disconnected systems and were removed. Making that distinction openly told experienced teams that their knowledge was respected, not overwritten.

Technology as the governing tool

Only then did technology come in, and never as the transformation itself. Its role was to make the new operating model consistent and visible in daily work.

Before the integration, each company ran its own small CRM and ERP systems, alongside offline spreadsheets and its own pricing principles and guidelines. We consolidated all of them into one platform. CRM became the single commercial record of who owned each account and how deal credit was shared. ERP became the transactional backbone, from quotation to invoice. Master data management gave every business the same definitions for customers, products, suppliers and margins.

A single platform only creates value if people use it, so adoption became our main measure of whether the change was working. It was not easy, and it did not happen by itself. It needed the visible endorsement of the former owners and senior executives. It needed training for middle managers and operating teams. And it needed recognition sessions and steady encouragement, so that the teams leading the way were acknowledged and the rest felt confident to follow.

Together, the platform and its adoption settled the questions that had fuelled the conflict:

The question people argued about	The standard answer
Who owns the customer?	The CRM customer hierarchy and a named account owner
Who gets credit for a deal?	Deal-team rules, agreed before quotation
Can this price be approved?	Margin, credit and risk thresholds built into the quotation workflow
Can we deliver what we promise?	Connected supplier, production, quality and logistics milestones
Which number is correct?	Shared dashboards using governed data and agreed KPI definitions

AI in pricing: how machine learning turned negotiation into decision

Pricing was the most political process of all. Salespeople quoted from local cost sheets, and a price that looked profitable on the day could lose money once supplier costs, inventory or freight moved.

We built a pricing engine on machine learning (ML). It combined CRM, ERP, procurement, inventory and finance data, and learned from historical, current and forecast data to recommend a price range, expected margin, confidence level and risk flags. It calculated cost and transfer pricing across each cost and profit centre, estimated how costs would move by the delivery date given current inventory, and drew on the treasury team's FX forecasts and the win/loss ratios in the CRM pipeline. Because inventory was costed on a first-in, first-out (FIFO) basis, it could show, for example, that a quote priced for a 12% margin held only if the order drew on existing, lower-cost stock. If the project needed new material bought at current prices, the same quote would deliver closer to 7%.

The model then projected each deal's effect on the top and bottom lines, mapped against quarter-to-date and year-to-date performance in the financial plan, and recommended which deals to prioritise and how to meet the commercial objective, whether sales target, profitability or market share. With US–China trade tensions rising and the global economy uncertain, that gave senior executives a clearer way to manage tariff, FX, interest-rate, credit and trading risk before committing to a price.

The engine did not decide. It gave management three visible choices: grow sales with a sharper price; protect profit by changing the price, specification or terms; or buy market share deliberately, as a recorded investment with a clear case for future revenue. A negotiation that used to depend on who argued hardest became a decision anyone could see and explain.

How the model was introduced mattered as much as the model itself. It was positioned as decision support, not a replacement for commercial judgement, and it came with the same communication, training and adoption plan as every other part of the change. That is why the sales teams used it rather than worked around it.

Its biggest effect, though, was on how leaders worked together. Senior executives could now review the important deals side by side and decide collectively which to pursue to protect the core business and gain share. Resource allocation became a decision about the whole business rather than a contest between former companies. The single source of truth had become a single source of trust.

The results

After the 12-month transformation programme:

Measure	Result
Sales	+11%, finishing 6.7% above the planned sales figure
Profit	+16%
Operating costs	−14%
Operating efficiency	+17%
Customer and supplier network	One digital flow of proposals, events, purchase orders, acknowledgements and invoices across 1,000+ customers and suppliers

The change I value most does not appear in the table. By my own estimate, leadership meetings became more than twice as effective. Instead of questioning one another's definitions and standards, senior executives spent their time on strategy. The question in the room changed from “Why has your team caused this problem?” to “What does the data say, and which decision is best for the customer and the P&L?”

Seven legacy businesses had become one team, and the network of customers and suppliers around them had taken its first step towards a digitalised supply chain.

THE FRAMEWORK: DATA AS THE REFEREE

Most resistance to change comes down to five questions people ask, often silently. Leaders who answer them with respect, transparent rules and shared data, rather than authority alone, take the politics out of change.

The people question	Why it creates resistance	How it was answered here
1. Is there a place for me, and are my interests understood?	Uncertainty about role, status and personal commitments, such as earn-outs	Respect for expertise; interests put on the table; a plan that worked for every party
2. Whose customer, whose credit?	Fear of losing accounts and recognition	CRM account ownership and deal-credit rules agreed before quotation
3. Which number is true?	Competing spreadsheets turn data into a weapon	Master data, governed dashboards and agreed KPI definitions
4. Will the system decide for me?	Fear that technology removes judgement	AI recommends and flags risks; people make, and own, the decision
5. What am I measured on?	Local targets reward local behaviour	Full KPI visibility and shared measures alongside local P&L

The order matters. Respect first, then fair rules, then a shared truth, then trust in the tools, then measures that hold it all in place. Skip the first and even the best technology becomes another battleground.

The same five questions apply to any AI rollout today. Whether you are introducing an ML pricing model or AI agents into finance, procurement or customer service, people will ask whether there is a place for them, who gets the credit, which number is true, whether the system will decide for them, and what they will be measured on. Answer those first, and adoption follows.

WHAT THIS MEANS FOR YOUR LEADERSHIP TEAM

For CEOs and boards: align the deal with the integration

An acquisition is complete on paper long before it is complete in practice. If the earn-outs and performance commitments of acquired leaders pull against the integrated business, expect resistance, however good the plan. Work with the holding company and deal teams to align those incentives early, and ask one test question: are leadership meetings spent on strategy, or on arbitrating ownership?

For CFOs and P&L owners: make one number the only number

Agree definitions for revenue recognition, cost allocation and margin before you debate performance. Governed data shortens reconciliation and makes profitability comparable across units. Then use it to look forward: AI pricing shows the margin risk in a quote before the price reaches the client, not after the invoice.

For CHROs and people leaders: design the change for each level

One communication plan does not fit an organisation. Senior executives need to agree; middle managers need to see what is in it for them and how they will be measured; operating teams need the reasons, the training and a realistic pace of adoption. People accept a hard change faster than an unclear one.

For chief commercial officers and sales leaders: replace opinion with evidence

Agree account ownership and deal credit before the quote, not after the win. Sales teams will trust a pricing engine that explains its recommendation and leaves them the final call.

For CIOs, CDOs and technology leaders: build the guardrail, not the goal

CRM, ERP and master data management are not the transformation. They are how the operating model shows up in daily work. Design them around the decisions people argue about most, and treat data quality as a shared business KPI rather than an IT metric. When you introduce AI, make it explainable: a recommendation people can understand is a recommendation they will use.

THE LAST WORD

Technology does not transform an organisation; its people do. What technology can give them is a single version of the truth, and that is the foundation on which trust, alignment and lasting change are built.

In your next leadership meeting, how much time will be spent deciding which number is right, and how much deciding what to do about it?

—

Until next week,

Artificial Ian

→ Found this useful? Forward it to a leader in the middle of an integration.

→ Subscribe to The AI Executive Briefing, weekly in your LinkedIn inbox.

Sources: Gartner, “Gartner HR Research Finds Less Than Half of Employees Achieved the Change Goals Set by Their Organization”, press release, 9 October 2025 (surveys of 2,500+ employees and nearly 1,000 leaders, April 2025). Bain & Company, M&A Report 2023, “Cultural Integration” (M&A Practitioners' 2023 Outlook Survey). McKinsey & Company, “B2B pricing: Navigating the next phase of the AI revolution”, 7 April 2026 (Agentic AI in Pricing Survey, November 2025, n = 419). Integration results: the author's own programme data.