How to automate management reporting without replacing your existing systems

90% of finance teams that already bought planning software still run Excel alongside it, and that figure has not moved in a year. Three routes to a shorter cycle, which step each one actually fixes, and the order of work that collapses the days.

The pack takes too long, and the obvious answer is to buy something. Before that, one piece of evidence worth sitting with, because it comes from the vendors themselves.

What the platform purchase did not remove

What buying the platform did not remove

These are finance teams that already own planning software, reported by a company that sells planning software.

• Still rely on Excel alongside their planning platform
• Spreadsheets are the most-used tool for budgeting and forecasting
• Only moderate or limited integration between FP&A tools and source systems
• Rank spreadsheet reliance as their number one technology challenge

Source: Vena 2026 FP&A Impact Report, administered October 2025 with Benchmarkit, n=431 finance professionals from under $50M to above $1B revenue across North America, Europe and the rest of the world. Vendor-commissioned, and the finding runs against the vendor's interest, which is why it is worth reading.

Vena's 2026 FP&A Impact Report, administered in October 2025 with the research firm Benchmarkit across 431 finance professionals at organisations from under $50M to above $1B in revenue, looked specifically at teams that already own planning software. It found 90% still relying on Excel alongside that platform for at least some models and reports. The year before it was 89%, so the number has not moved.

Spreadsheets were the most-used tool for budgeting and forecasting, cited by 61%. A third ranked spreadsheet reliance as their number one technology challenge, and 66% put it in their top three. On integration, 51% reported only moderate or limited connectivity between their FP&A tools and their source systems, and 7% had none at all and depended on manual uploads.

A second survey, run by Centiment for Vena between 16 and 18 July 2026 across 364 finance professionals from Finance Manager to CFO at organisations above 200 employees in the US, Canada and the UK, found 92% of planning-software owners still using Excel at least weekly, and 53% of that group expecting their Excel reliance to increase over the next year against 8% expecting it to fall.

Both are vendor-commissioned and both run against the vendor's commercial interest, which is the reason to take them seriously rather than discount them.

The conclusion is not that planning platforms are useless. It is that the platform addresses the assembly step, and in most cycles the assembly step is not where the days are.

Three routes, and what each actually fixes

Three routes, and what each one actually fixes

Pick on the basis of which step in your cycle holds the hours. Not on which route sounds most modern.

A quarter of surveyed companies above $1B in revenue still run Excel as their primary planning tool. Scale is not what forces route three.

Source: Vena 2026 FP&A Impact Report, administered October 2025 with Benchmarkit, n=431; Thinklytics engagement pattern across the 14 automated executive reporting engagements in the case library.

Certify the metrics and automate assembly on the systems you have. Fixes reconciliation, exception chasing, assembly and rework, which is four of the six steps. Leaves you on a dated stack, which is fine until something goes end of life.

Add a reporting or semantic layer above the existing systems. Does the same, and adds one place to query, which is what removes the ad hoc request queue. Leaves the source systems disagreeing at the record level, so if the problem is duplicate customers rather than conflicting rules, this is not the route.

Replace the source platform. Fixes the export problem and the end-of-life problem. Carries every definition dispute forward into the new system unless someone explicitly stops it, which is the failure mode worth planning against.

Choose on the basis of which cycle step holds your hours. That measurement is covered in why monthly reporting takes so much manual work, and it takes one reporting cycle.

What forces route three, and what does not

Three conditions force a platform replacement, and company size is not one of them.

A system at end of life, with no vendor support and no patch path. An export that cannot be scheduled, usually a screen-scrape or a report that only runs interactively. Or a source that cannot be integrated at all, which in the Vena data applies to the 7% depending on manual uploads.

Against that, nearly a quarter of surveyed companies above $1B in annual revenue said they rely on Excel as their primary planning tool. Scale does not force the migration. Assuming it does is how a large organisation funds a 26-week consolidation for a problem that was eleven weeks of definition work.

Our own position on this is on the record in why we rarely recommend platform migration.

The order of work

The order that collapses the cycle

Each step is cheap and makes the next one cheaper. Running them in reverse is the most common way an automation project overruns.

• Time the six steps separately
• Certify the metrics that get argued about
• Automate the extract and the exception report
• Bind the pack template to the certified numbers
• Only then ask whether a system needs replacing

Three weeks to two days at a regional bank came from defining all 140 call report line items and automating extraction and calculation. The platform stayed.

Source: Thinklytics engagement pattern across the 14 automated executive reporting engagements in the case library.

Five steps, in this order, because each one is cheap and makes the next one cheaper.

Time the six cycle steps separately. One cycle. No software decisions until this is done.

Certify the metrics that get argued about. Not every metric. The contested ones. One written definition each, with a named owner, answering when it counts, one row per what, what is in and out, what gets subtracted, and at what rate.

Automate the extract, and add an exception report that runs before the pack. The exception report is the step most often skipped and it carries a lot of the benefit. It fails loudly on the things that historically caused rework: a late journal, a missing cost centre, an entity on its own chart of accounts, a balance that moved after sign-off. Found before assembly it is a half-hour conversation. Found by a reviewer it is three days.

Bind the pack template to the certified numbers. This removes the reformatting and most of the remaining rework, because a number changing upstream no longer means rebuilding four slides.

Only then ask whether a system needs replacing. By this point you know, because the steps that are still slow are the ones the first four could not reach.

Commentary stays human throughout. It is the analysis, and it is what the recovered hours are for. Generated commentary changes the writing rather than the cycle, and it adds a verification obligation: Workiva's MidYear Executive Benchmark Survey, run by Ascend2 and fielded in May 2026 across 2,272 finance, risk, sustainability and legal professionals, found 26% saying their own internal AI audits had detected AI errors that reached external audiences or the board.

What this has delivered in practice

A regional bank cut quarterly call report preparation from three weeks to two days. The work was defining all 140 call report line items precisely, automating extraction and calculation, and adding a validation step that flagged data quality issues before submission. Annual compliance labour fell $680K, four senior analysts went back to finance work, and the next regulatory exam produced no matters requiring attention. The platform stayed. See the regional bank governance engagement.

An insurer took statutory NAIC filing preparation from six weeks to four days, and from 480 person-hours a quarter to 32, with zero restatements across the first four quarters after deployment. Kaiser Permanente consolidated 14 regional definitions of a patient encounter onto the enterprise data warehouse it already had, in 11 weeks, with clinical reporting running without interruption throughout, and reconciliation labour falling from $2.1M a year to under $180K. See the Kaiser Permanente metric governance engagement.

Where a platform consolidation was the right answer, it took longer and the shared semantic layer was built as part of the migration rather than after it. A national telecom replaced nine reporting systems over 26 weeks, cut the monthly close from 18 days to three, and saved $3.1M a year in infrastructure. That is a real outcome and a different size of commitment.

What we would do first

Measure, then certify, then automate. In that order, with no procurement decision until the first measurement exists.

If the hours turn out to sit in reconciliation, the delivery surfaces are AI reporting automation for the assembly and analytics and BI for the layer above it, with agentic BI implementation and decision support systems where the executive audience needs to ask follow-up questions without raising a request. On Power BI specifically, Power BI consulting; on SAP estates, SAP reporting and analytics modernisation.

To put a number on either route before asking for budget, the reporting baseline and benefits worksheet is what we use with clients, and how to scope an executive reporting automation project covers what belongs in the statement of work. The full set of work in this area sits under reporting takes too much work.

Frequently asked questions

Can we automate management reporting without replacing our systems?

In most cases yes, because the step holding the hours is usually reconciliation rather than extraction, and reconciliation is fixed by writing definitions down and moving exception handling in front of the pack. A regional bank cut quarterly call report preparation from three weeks to two days by defining all 140 line items and automating extraction and calculation, with the existing platform left in place. Kaiser Permanente consolidated 14 regional definitions on the enterprise data warehouse it already had.

What are the three routes and how do I choose?

Certify the metrics and automate assembly on the systems you have; add a reporting or semantic layer above them; or replace the source platform. Choose on which of the six cycle steps holds your hours. If it is reconciliation and rework, route one. If it is also the ad hoc request queue, route two adds one place to query. Route three is for an end-of-life system or an export that cannot be scheduled, and it carries every definition dispute forward into the new system unless someone stops it.

Does buying a planning platform remove the manual work?

Not on the available evidence, and the evidence comes from the vendors. Vena's 2026 FP&A Impact Report, administered October 2025 with Benchmarkit across 431 finance professionals, found 90% still relying on Excel alongside their planning platform, unchanged from 89% a year earlier, with 51% reporting only moderate or limited integration between their FP&A tools and source systems and 7% with none at all. A separate Centiment survey for Vena, fielded 16 to 18 July 2026 across 364 finance professionals, found 92% of planning-software owners still using Excel weekly and 53% expecting that reliance to increase.

What order should the work happen in?

Time the six cycle steps separately, certify the metrics that get argued about, automate the extract plus an exception report that runs before the pack, bind the pack template to the certified numbers, and only then ask whether a system needs replacing. Each step is cheap and makes the next one cheaper. Running it in reverse, with the platform first, is the most common way one of these projects overruns, because a migration carries the existing definitions forward and then gets blamed for not fixing them.

What is an exception report and why does it come first?

A check that runs before the pack is assembled and fails loudly on the things that historically cause rework: a late journal, a missing cost centre, an entity on a different chart of accounts, a balance that moved after sign-off. Running it after the pack means the exception is found by a reviewer, which is the expensive path. Running it first turns a three-day rework loop into a half-hour conversation before anyone has built a slide.

Is company size what forces a platform replacement?

No. In the Vena data, nearly a quarter of respondents above $1B in annual revenue said they rely on Excel as their primary planning tool. What forces route three is an end-of-life system, an export that cannot be scheduled, or a source that cannot be integrated at all. Scale on its own does not, and assuming it does is how a large company ends up funding a migration for a reconciliation problem.

What about AI-generated commentary?

It changes the writing, not the cycle time, and it should be kept out of the first phase. The commentary step is the analysis and the one worth protecting. Generating it from numbers a person assembled by hand leaves every manual step in place while adding a verification obligation. Workiva's MidYear Executive Benchmark Survey, run by Ascend2 and fielded May 2026 across 2,272 finance, risk, sustainability and legal professionals, found 26% saying their own internal AI audits detected errors that had reached external audiences or the board.

How long does the no-new-systems route take?

In our case library the engagements that shortened a reporting cycle on the platform already in place ran 8 to 20 weeks, and the duration tracked the number of contested numbers in the pack rather than the size of the company. Eight weeks for five ARR definitions. Eleven weeks for 14 regional definitions, and eleven for a statutory filing with a defined line-item set. Twenty weeks for 42 metrics across eight administrative units. Platform consolidations in the same library ran 16 to 26 weeks.

The work behind this

Fourteen engagements in the case library carry automated executive reporting. The ones that left the platform in place ran 8 to 20 weeks; the consolidations ran 16 to 26. Each states which route was taken and what the cycle time became.

Automated executive reporting, 14 engagements.