Precanto co-founders and product leaders went on a listening tour the last couple of months. 

Are we building a P&L management platform with an “outside in” approach that will meet our customers’ expectations? How should Precanto leverage automation and data science to deliver real time visibility and predictions for a company’s largest spend category, headcount and related personnel costs? How do we scale what current legacy planning solutions cannot do with data science? What product roadmap next generation functionality will appeal to our customers and prospects?

We conducted more than 100 interviews with CFOs, FP&A executives, talent acquisition leaders, and budget owners from large publicly traded companies to hyper-growth start-ups across a variety of industries and regions. So, what have we learned?

Forecasting headcount and related personnel costs is challenging and costly, if you get it wrong

I will not lie. There were some folks that said “headcount doesn’t matter” or “headcount doesn’t fluctuate.” (I may have bitten my tongue.) “How can 60-85% of expenses not matter and never fluctuate?” COVID, hybrid workforce, layoffs, reorganizations, acquisitions … Fortunately, those folks were in the minority and in fairness, the more I thought about this, these folks may not have had a front row seat to appreciate the gymnastics their FP&A teams undertake to keep the Titanic afloat. 

Headcount spend forecasting is …. VERY challenging, tedious, and costly, especially if you get it wrong.

A CFO of a 1,500 FTEs company made a poignant observation, “In my experience, headcount spend is always million dollar errors often going in either direction and sometimes you get lucky and they cancel each other out. Often though, you don’t get lucky. If you’re publicly traded, you can’t afford to be lucky. And if you’re privately held, you want to keep your job and not have missed too many forecasts.” 

‍

When asked “Why is headcount forecasting challenging”, here is what we heard:

• There is too much transactional headcount data to quickly activate insights and predictions
• There are many variables for headcount forecasting (e.g., start date, location, level, dept, etc.)
• Despite significant investments in our legacy planning solution, teams rely on spreadsheets
• Spreadsheets cannot handle large volumes of data from HR and accounting systems
• Budget owners and talent acquisition teams do not have access to planning solutions
• Finance teams to not have access to all HR and recruiting systems
• Only FP&A or the systems admin understand the human built formulas for the legacy solution
• There is no “single source of truth” to perform real time scenario plans for headcount decisions
  ‍

When asked “Why is it costly to get headcount wrong”, here is what we heard:

• We missed earnings and our market cap dropped, solely because we missed on headcount
• We did multiple, significant layoffs using spreadsheets. We don’t know if we got it right
• We let go 12% of our workforce. We don’t know how we can run our business beyond this RIF
• We make bad decisions (real estate, benefits, bonus plans) because there is no spend visibility
• We have many TBHs not filled each quarter that impact our ability to hit OKRs and business goals
• We are hiring aggressively, but pretty sure we may not hit targets because we lack recruiters
• We have high FP&A attrition and poor employee satisfaction because this is too hard to manage
  ‍

Do It Yourself (“DIY”) data science doesn’t work for headcount expense predictions

Enterprises are investing in DIY with data science, just not in back office functions like accounting and finance. Many CFOs and FP&A leaders have some training in statistical reasoning, regression modeling, and predictive analytics, but to scale what they know would require significant investments and support from those that are more technical. And DIY is costly, time consuming, and prone to errors, especially if not leveraging best practices or understanding accounting nuances.

When asked “Have you considered DIY for headcount spend predictions”, here is what we heard:

• Our data scientists know data science. They don’t know accounting and finance. What’s GAAP?
• We have data scientists and they need to be 100% focused on product usage and customers
• Even if we built out a machine learning (“ML”) approach, our legacy planning solutions can’t handle 3+ years of historical data at granular level of detail that is needed (e.g., sub-ledger from accounting systems, transactional data from recruiting system)
• Training and maintaining ML models in-house with all that the FP&A team needs to do would be impossible in terms of time, costs, and best practices
• Our CFO wants fast time to value for headcount spend forecasting. DIY isn’t the answer

Enterprises are excited for automation and data science for expense forecasting

Thanks to companies like Aviso, BoostupAi, Clari, and Gong, enterprises have experienced the power of automation and data science for revenue forecasting. Enterprises of all sizes - less than 1K FTEs to those with more than 30K FTEs - are exploring how to bring automation and data science to the back office to help with expense visibility and predictions. Because headcount and related personnel spend is 60-85% of expenses, enterprises are eager to automate, scale, predict, scenario plan, and align decisions for their largest spend category. If AI/ML worked for revenue forecasting, why not the rest of the P&L?
‍

When asked “What would be your data science nirvana for headcount spend”, here is what we heard:

• Automation to help clean up our applicant tracking system data to power ML models
• Real time headcount spend visibility that simulates a daily close for all headcount costs
• Data science predictions across both direct (base, bonus, payroll tax, benefits) and indirect (real estate, software, travel and entertainment) spend categories 
• Activated headcount insights that challenges our human-built formulas and budget drivers
• Collaborative solution for FP&A, talent acquisition, and business partners that has built in finance context so non-finance folks don’t need to learn how to model their headcount decisions
• P&L platform that predicts headcount productivity by function
• CFO and FP&A executive alerts to help teams course correct throughout the quarter, without waiting on accounting close
• Dynamic scenario planning to answer a CFO’s question “what headcount do I need in what region, function, and by level 3 years from now if I assume these revenue assumptions”
  ‍

Stay tuned for more updates from Precanto’s listening tour.

Are you willing to be interviewed by Precanto to help advance our product roadmap and improve our messaging and value proposition? Find me on LinkedIn or via Contact Us on our Precanto website.

PS - Shout out to all the CFOs, FP&A executives, talent acquisition leaders, and budget owners that took time out of their busy days to help validate (and challenge!) what Precanto is building. Thank you!