.png)
The algorithm served me up an old Steve Jobs interview this week. In his early Apple days he asked the accountants how they worked out what a product cost to make. They told him they used a standard cost, essentially an educated guess, and then adjusted for the difference between the guess and reality later on. He kept asking why, round after round, and eventually got to the honest answer, nobody actually knew the real cost, so everyone guessed and corrected afterwards. Every company did it, so it seemed normal.
Obviously, being Steve Jobs… he didn’t like what he heard. He insisted that costing accuracy could be 100% accurate with the right discipline and systems.
The question I am asking today is “Is Flash Reporting a concept that Jobs would have approved of?”.
What is Flash Reporting?
For those not working in Financial Reporting, let me first set the scene of what a Flash report is, else the rest of this isn’t going to mean a lot to you.
Each month, finance teams close the books on trading for the prior month, they make sure that all the numbers are correct and do accounting-y things that only need to happen once per month (like calculating depreciation on assets – no need to do this daily). This is called “the close”. It’s both boring and necessary.
Anyway, while this close activity is happening, business leaders want to know how the business performed this month so that newer, better decisions can be made and strategy can be adjusted accordingly. They’re waiting for the numbers. The close might take more than a whole week to execute.
Stakeholders don’t like that. Please imagine a person drumming their fingers on the table, hitting “Send / Receive” in their emails and sighing loudly.
To avoid finger-drumming, Finance invented “The Flash”. I think DC are lawyering-up. The Flash Report is basically a first cut of the numbers before the Close is 100% finished and agreed. I’s dotted and t’s crossed. The idea is that we’re offering insights with a caveat. “This is probbbbably right”. You get the idea.
Does my Steve Jobs comparison make sense yet?
What happens next?
After the close is 100% done, the Finance team will send a more complete reporting package, which is final, is trusted and is packed with insights. Companies tend to call this a “CFO Pack” or a “Board Report”, basically depending on how important it is deemed.
My first posting in consulting was designing and automating various reporting packs which has specific target audiences (End Markets, Business Units, Regions and the Group) that were released every month end. It was fascinating and taught me that everyone has an opinion of what’s the most important KPI, and invariably, no one agrees with each other.
If you take a step back you notice that a Board Pack is basically an extended Flash Report. It’s the same stuff, just a little bit more detailed and a little bit more confident.
Recap so far - two packs, potentially double-effort in production and review, which may or may not agree with each other.
Yes, that sound is Steve Jobs’ blood boiling.
But, the gap between the two reports is bigger than you’d expect. APQC's benchmarking puts the median time to complete the monthly close at six days, with some organisations taking ten or more. That's a long time to run the business on a number we've told everyone not to rely on.
Are decisions made on Flash Reports?
That’s the key question isn’t it? Do execs make decisions and change behaviour based on the report with the caveat? If no, could they just wait for the full report? If yes, if they’re wrong, where does the fault lie?
The probable and predictable answer is, it depends. Classic consulting.
Here are three theories. ONE – more repetitive, predictable businesses (think ARR) probably are pretty happy with the flash. TWO – it probably comes down to the individual reading the report. Do they have a propensity for risk? Do they sense any risk? THREE – depends on the risk profile of the business. I bet big-Pharma is less interested in flash reporting than play-doh manufactures.
How does VantagePoint treat Flash Reporting?
No, I’m not talking about our opinion, I’m talking about our internal reporting capability. A look behind the curtain if you will.
We’re a services and software-based business, so that means the majority of our revenue is variable and event based. Therefore, we like data. We like predicting things and we like regular updates.
We don’t have a flash report. Instead, we have a daily view of Actualised & Forecasted revenue based on our internal timesheeting process. We have live dashboards that can show expected revenue by consultant, by project against a plan which is maintained daily.
This is the only time that I’ll use VantagePoint and Steve Jobs in the same sentence: I think he’d like our approach.
Nonetheless, back on topic, in the AI-Era….
Can Flash Reports go the way of the dinosaurs?
Data integration used to be a months-long IT project, and now it can often be configured in days. APIs, MCPs and AI tools let us connect bespoke solutions like Lego, each piece small, all of them clicking together. This is a bit of a game changer for reducing time-to-insight and the cost of getting there.
Four ingredients to retire the flash
- Automated month-end journals. Remember the close? Depreciation, amortisation, lease accounting, accruals, prepayments. They’re all predictable & rule-based. They can be built in AI tools or subledgers, and if a journal can be predicted, it shouldn't need someone to post it by hand. Cut that bit out and you’ve just saved 1 day from getting your Board Pack to the Board.
- An integrated reporting framework. Think of it as a maths function, numbers go in and reports come out. Whether that's Workiva-style linked presentations or Power BI, the test is simple, can you put numbers in one place and have every table and chart update? Commentary follows the same logic. My rough estimate is that around 70% of board commentary is automatable, because "product X drove the biggest swing in sales between March and April" is really a ranking calculation with the right data at the right granularity. The rest needs institutional knowledge, so a good pack is designed for both, automated insight that drills from entity to account to product, and clearly marked places where a human explains what the numbers can't. Compiling the full-board report might have taken 12 working hours in Excel and PowerPoint. Take that time back!
- Protected time to review, in one place. AI does not mean that thinking is not necessary. Far from it. Finance teams still need to get together and chat. A one-hour calendar block on working day two to comb through the numbers together isn't avoidable, even in the age of AI, and it shouldn't be! That hour is where the team discusses and understands the results. Anyone who has worked with AI for a while will know your proximity to the detail fades a little with every generated output, so whether the numbers come from AI or a more traditional system, you need a forced check to validate what's been produced for you. Do it on a single call, just once, with everyone looking at the same numbers, and you lose the pass-around of files and the "v7_FINAL_final" that comes with it. It's also when the best ideas come up. Your Finance team may even enjoy it. “How can we make this better?”, they’ll ask.
- Agreed decision thresholds. Most finance teams already set materiality thoughtfully, but it's usually framed around reporting and audit risk, rather than the decisions downstream. To retire the flash, I think we need a second lens… how big would an error need to be to change a decision? The answer will differ by metric. A manufacturer tracking gross margin per thousand units might need that line right to the decimal point, while a small movement in group EBITDA might not change anything the board does. Each business has to set those thresholds, ideally agreed with the board, and then report against them. A track record helps too, so if the day-two pack has needed corrections in recent months, show how big they were, e.g. the average movement per line between day two and the final numbers over the last six months, or a tolerance band on key charts. Telling the board "this is where we've been wrong before, and by how much" builds more trust than a caveat ever will.
Is this realistic?
If a business already has analysis and decision-making forums in place, and a history of good data, deciding how to optimise this process is relatively straightforward. For businesses that are still scrambling, where people are under pressure just to get their tasks done, what I'm describing will probably feel utopian and a long way off.
That's where leaders need to step in and carve out time to get to the promised land, and accept that it takes some short-term investment in capacity. Without that, you'll start to fall behind, and your team will keep fielding the "why isn't this faster, surely AI can do it for you?" question, with growing resentment.
Jobs didn't accept "everyone guesses" as an answer. In 2026, I don't think Finance teams need to either.
What do you think? Would your board accept a day-two pack with a known tolerance? What's stopping your team?
If you'd like a clearer path to getting there, VantagePoint runs short, sharp consulting engagements that give finance teams a two-year roadmap to modernise their processes. If that sounds useful, message me at matt.roberts@vantagepointconsulting.com
