Only 10% of Companies Have Fully Integrated Planning. Apparently, the Other 90% Have Meetings.

Strategy has a plan.

Finance has a budget.

Operations has a staffing schedule, a production target, and three reasons the budget is wrong.

Everyone has a spreadsheet.

Congratulations. We have achieved alignment in the sense that everybody agrees there should probably be some.

New research gives this familiar situation an uncomfortable number. The 2026 FP&A Trends Insights Paper, published October 12, reports that just 10% of organizations fully integrate strategic, financial, and operational planning. Among that small group, 41% rate their FP&A function as optimized or performing well. Among organizations with no integration, the comparable figure is 11%.

That is a 3.7-times difference in the share reporting strong performance. It is an association in a survey, not proof that buying a planning platform makes your team 3.7 times better.

Still, it is a remarkable gap. And I think the most interesting question isn’t why 90% haven’t finished integrating planning. It’s why we keep treating integration as a software installation when so much of the problem happens before anyone opens the software.

What Integrated Planning Actually Means

Integrated planning is an unglamorous idea with expensive consequences: the strategy, operating decisions, and financial expectations should describe the same business.

If leadership wants to grow revenue by 15%, Sales needs a credible pipeline and conversion assumptions. Operations needs the capacity to deliver. HR needs the hiring plan. Procurement needs the supplier commitments. Finance needs the margin, working-capital and cash consequences. And all those assumptions need to change together when reality changes.

That last part is where the wheels tend to leave the vehicle.

A connected planning process is not merely a shared database. It is a system of shared drivers, definitions, timing, ownership, and decisions. If Sales changes its forecast and Operations learns about it at the monthly review, the two functions may be using the same application but they are not operating from the same plan.

I’ve always found the distinction useful: a number can be technically connected without being operationally agreed. The formula works. The meeting does not.

Integration also does not require eliminating departmental expertise. Sales should still understand its funnel better than Finance. Operations should understand constraints that never appear in a chart of accounts. The point is to make those different views compatible enough that management can see the trade-offs before committing resources.

The Survey Numbers Deserve a Closer Look

The headline 10% is striking, but I wouldn’t use it to declare that every other organization has failed. Full integration is a demanding standard. Some organizations have partial integration, and the maturity of a planning process can vary dramatically by business unit or decision type.

Another FP&A Trends discussion of extended planning breaks the population into roughly 10% fully integrated, 22% reporting no integration, and 68% somewhere between. It also notes that only 7% directly link strategy to operations. Those distinctions matter. The middle group isn’t necessarily doing nothing; it may be connecting Finance to Operations while leaving strategy in a separate conversation.

The latest research’s 41%-versus-11% comparison tells us that stronger integration and stronger self-reported FP&A performance travel together. It doesn’t establish causality. Better-managed organizations may be more likely to invest in integration, and better planning may reinforce better management. Both could be true.

Meanwhile, the Association for Financial Professionals’ 2026 benchmarking survey, based on 332 finance professionals in 54 countries, reports that only 38% use structured scenario planning. It also finds that structured scenario users complete budgets 11% faster on average, while the average budgeting cycle remains close to nine weeks.

Different surveys ask different questions, so I would not combine their percentages into one grand statistic. What they do suggest is a persistent distance between knowing what modern planning should look like and making it work during a real budget cycle.

Why Departments Can Agree in a Meeting and Disagree in the Forecast

Consider a hypothetical manufacturer preparing its next annual plan. The CEO wants 12% growth. Sales expects to deliver it through new accounts. Operations has enough capacity for only 7% more volume without overtime or additional equipment. Procurement expects material prices to rise. Finance’s preliminary plan shows margin improvement.

Every one of those inputs may be reasonable in isolation. Together, they might describe four different companies.

The budget meeting ends with agreement on the revenue target. The capacity decision gets postponed. Procurement’s price assumptions live in another workbook. And the margin improvement survives because nobody has yet forced all four assumptions into the same model.

Three months later, management asks why revenue is ahead of plan but operating profit is behind. The answer is not necessarily a forecasting error. It may be an integration error that was baked into the plan before the year began.

I’ve seen how easy it is for a planning process to reward agreement over clarity. Everyone wants to leave the meeting with an approved number. The uncomfortable questions can wait. Unfortunately, the business usually does not wait with them.

And here’s the part I wish more budgeting meetings would embrace: disagreement is useful. A capacity constraint is useful. An uncomfortable cash forecast is useful. Discovering those things after the budget is approved is less useful. The purpose of integrated planning isn’t to eliminate disagreement. It’s to have it while management can still do something about it.

Forget the Integration Diagram. Show Me the Decision.

This is the part of the FP&A Trends paper that I find most useful: start with an important decision and trace backward to the layer that limits its quality. That is a more practical starting point than announcing a companywide transformation and scheduling seventeen discovery workshops.

Take a hiring decision. A business leader wants six new account managers. Before Finance approves the cost, it needs to understand expected bookings, onboarding time, productivity ramp, customer demand, territory capacity, compensation, and cash timing.

Which assumptions are owned? Which are shared? Which can change? Who has authority to approve a different hiring pace when bookings miss the target?

If those answers are unclear, the problem is not that the hiring forecast needs a prettier dashboard. The problem is that the decision has no reliable operating agreement behind it.

I would pick one decision that matters, map its drivers and owners, and make the process work end to end. Then expand to the next decision. It is less glamorous than a transformation roadmap with an ambitious acronym. It also gives the organization a chance to learn something before spending the entire budget.

Before We Buy Software, Can We Ask the Humans What They’re Doing?

Finance teams often inherit planning processes built around departmental habits. A department head may think of a budget as permission to spend. Finance may think of it as a financial forecast. Leadership may think of it as a performance commitment. Those are not interchangeable definitions.

Ask three people what an approved budget means, and you may get four answers. One person will probably say the fourth answer is in an email.

Before changing systems, I’d sit with the people who create the operating inputs. What decisions do they make each week? Which numbers do they trust? What do they know that the financial model misses? What assumptions change first when customers behave differently?

These conversations aren’t a soft-skills detour from planning. They are how Finance discovers the operating logic it needs to model.

It also changes the relationship. Instead of arriving with a spreadsheet and asking people to populate the yellow cells, Finance can explain how their knowledge changes the decisions management makes. That is a far more persuasive reason to participate.

And yes, eventually there may be yellow cells. We’re Finance. We have traditions.

A Practical Example: When a Growth Target Meets Capacity

Let’s put numbers around the problem. The following scenario is illustrative, not a reported company result.

Assume a business has $100 million in annual revenue and targets 12% growth. Its current gross margin is 40%. The Sales plan assumes the extra $12 million can be delivered with the existing cost structure. On that basis, the incremental gross profit appears to be $4.8 million.

Operations says capacity supports only $7 million of additional sales at the current cost structure. The remaining $5 million would require overtime and expedited sourcing, reducing its gross margin to 25%.

Now the economics look different. The first $7 million produces $2.8 million of gross profit. The next $5 million produces $1.25 million. Total incremental gross profit is $4.05 million, not $4.8 million.

That’s a $750,000 difference before considering additional working capital, equipment or staffing.

We didn’t need an algorithm to find the missing $750,000. We needed Sales and Operations to have the same conversation before Finance calculated the margin. I realize that sounds less exciting than predictive analytics. It’s also a rather good way to avoid explaining a surprise to the board.

Now imagine leadership can invest $1 million in equipment that raises capacity and restores part of the lost margin. Suddenly the planning process becomes a decision about payback, cash, execution risk and timing rather than an argument about which department submitted the right forecast.

That is what integration should make possible: not merely a consolidated number, but a choice management can evaluate.

Five Conversations I’d Have Before Buying Another Planning System

First: agree on what we’re deciding. Are we trying to maximize growth, protect cash, improve margin or preserve service levels? Those objectives can conflict. Someone needs to say which one wins under which conditions.

Second: agree on the drivers before the graphs disagree. A sales forecast, production plan and staffing model should use compatible volume, timing and productivity assumptions. They do not need identical levels of detail, but their relationships must be explicit.

Third: stop making Finance the owner of everybody else’s assumptions. Each material driver needs a person responsible for the input and a process for changing it. Finance should not quietly become the owner of every number simply because the numbers eventually appear in Finance’s report.

Fourth: decide when we actually do something. If bookings fall 8% below plan, what happens? Do we slow hiring, revise purchasing, run a downside scenario, or wait until month-end? A threshold without an owner is a very organized way to remain surprised.

Fifth: write down what we decided, before everyone remembers it differently. Record what changed, why it changed, who approved the response and what financial impact was expected. When the next forecast arrives, the team can evaluate the decision rather than reconstruct the conversation from memory.

These agreements are less exciting than a product demonstration. They are also the things that make the product demonstration relevant to the business.

The Software Is Not Going to Have the Awkward Conversation for You

Before anyone accuses me of wanting to run a multinational business on one heroic spreadsheet and a good attitude, I like planning technology. Good tools connect data, automate calculations, preserve versions and make scenarios possible while the question is still relevant.

What I don’t like is asking software to settle an argument management hasn’t finished having.

But technology cannot decide whether Sales owns pipeline conversion or Finance owns the revenue commitment. It cannot resolve a compensation plan that rewards volume while the corporate strategy prioritizes margin. And it cannot make department heads trust a forecast they were never invited to shape.

The 2026 FP&A Trends survey discussion hosted by OneStream highlights the foundational challenge: 81% of teams use data in most decisions, but only 19% rate their data quality as advanced, while 47% of FP&A time still goes toward data collection. The survey also finds that integrated planning remains rare.

That combination should make any CFO pause before assuming that adding AI will fix planning. AI can accelerate analysis. It can also accelerate the circulation of inconsistent assumptions.

I’d evaluate planning technology against actual decisions: Can a business leader see the cash consequence of a changed sales forecast? Can Operations update a capacity constraint without breaking Finance’s model? Can we compare scenarios with consistent definitions? Can we identify which assumption changed the answer?

If the answer is yes, technology is earning its place. If the answer is merely that the dashboard refreshes every hour, wonderful. We can now be confused in near real time. I have a few follow-up questions.

Give Me 90 Days and One Decision That Keeps Going Sideways

If someone walked into my office with a 90-day plan to integrate every budget, forecast, strategic target and operating metric, I’d offer them a chair. They sound like they’ll need one.

I’d rather spend those 90 days proving we can improve one consequential decision. Hiring. Inventory. Capacity. Pricing. Something with enough dollars attached that management will notice when it improves. The timeline is illustrative, and a complex organization may need longer. I’d be willing to change it as we learn. That’s rather the point of planning.

Days 1–15: choose the decision and map the people. Identify the decision owner, Finance partner, operating contributors, data sources and current approval path. Interview the people who actually make the inputs. Document where definitions conflict.

Days 16–30: define the drivers. Agree on the handful of variables that move the outcome. Establish units, timing, assumptions, owners and acceptable ranges. Identify the points at which an operating change must become a financial change.

Days 31–50: build a connected model. Use existing systems where possible. Link the operating assumptions to revenue, costs, working capital and cash. Build at least a base case and two decision-relevant alternatives. Avoid spending the entire pilot making the interface beautiful.

Days 51–70: test a real decision. Run the model with the business. Introduce an unexpected change. Ask whether the team can update assumptions, understand the trade-offs and identify who must approve the response.

Days 71–90: measure and expand. Compare the process with the previous approach. How long did the decision take? How many reconciliations were needed? Were important assumptions identified earlier? Did leaders change an action because they could see its financial consequences?

The last question is my favorite. A faster forecast is useful. A better decision is the reason we’re doing this.

Please Don’t Tell Me How Many Dashboards You Launched

I’ve sat through enough systems discussions to know a dashboard can become the star of the meeting. Everyone admires the colors. Someone asks whether it refreshes automatically. Finance smiles politely while wondering if the numbers agree with Operations. I like a good dashboard. I just don’t think it deserves a standing ovation for displaying a disagreement faster.

Integration projects report technical progress: data feeds completed, users onboarded, dashboards launched. Fine. Those are milestones. They aren’t the business result.

I’d track decision latency—the time between identifying a material change and agreeing on a response. I’d measure forecast reconciliation effort, the frequency of conflicting assumptions, the percentage of material drivers with named owners, and the number of decisions evaluated through shared scenarios.

Here’s the test I’d care about most: did anyone make a different decision? Not a nicer slide. Not a faster email. A different decision, made earlier, with better information.

I’d also look for evidence that the organization acts differently. Did the hiring plan adjust earlier? Did purchasing avoid a capacity mismatch? Did management understand the cash impact of a growth decision before approving it?

Not every improvement can be assigned a clean dollar value. We should resist inventing one. But we can often document a credible chain from better information to a changed decision to a measurable outcome.

And if the team spends less time arguing about which number is correct, that is not a trivial benefit. It means people can finally spend more time discussing what to do about the number.

The Real Problem Isn’t the Missing Platform

Ten percent. That’s the number I keep coming back to.

Not because the other 90% are hopeless. I suspect many have smart people, serious budgets and systems expensive enough that nobody wants to admit the assumptions still don’t agree.

The finding is a wake-up call, but I don’t think the answer is to tell the other 90% to buy software immediately.

Start with the decision that keeps going wrong. Find the operating assumption Finance learns about too late. Identify the department that has a perfectly reasonable plan which conflicts with everybody else’s perfectly reasonable plan.

Then connect the people, drivers and financial consequences around that decision.

When those connections work, a planning platform can make them faster, more scalable and more reliable. When they don’t work, the platform can make the disagreement look wonderfully professional.

We don’t need more beautiful versions of three different plans.

We need one business making decisions with its eyes open.