Improved Revenue Forecasting

Paramount among the many analytic challenges facing Law firm CFOs and their financial staff is accurately forecasting cash. Relying on the law of large numbers, most firms assume that prior averages will hold, so they use history-based, firm-wide performance ratios to obtain cash flow projections.

Simplistic Forecasting

Accurate information is essential for effective forecasting, however, and historically many firms have taken a simplistic approach to getting at this information:

• They determine work value by timekeeper (number of hours x standard rate).
• Then they apply realizations (amount billed against standard, amount collected against amount billed) to determine the collectable work value.
• At this point, a number of variations on essentially the same approach are used. Prominent among these is to spread this collectable value across the previous year's collection pattern to derive budgeted collections.

For example if we budget $100 million of work value at standard, historically bill 96% of standard, and historically collect 98% of billings, then we will budget collections of ($100 million x .96 x .98) = $94 million. If in January of this year we obtained 10% of our annual collections, we will budget collections for January of next year at $9.4 million.

In a growing firm, this simplistic approach has an inherent weakness: over-budgeting revenue. The problem is that not all of this year's growing production will be collected this year. It takes several months to collect on work, so that some of the increased revenue will be booked but not available as cash until the following year. Moreover, firm growth usually occurs gradually over a year, so that the firm shouldn't expect to see much of the total expected annual revenue increment in the earlier months of the year being budgeted. It's not these patterns themselves that cause the problem, but that firms aren't accounting for them in their shortcut forecasting process.

Many firms utilize a historic haircut to right this wrong. For example, if the firm is growing at a 10% annual growth rate, they will reduce the annual budgeted collections by up to 10%. This approach previously worked just fine for firms, but is now problematic for three reasons:

• In today's consolidating and competitive environment, the formerly reasonable assumption of steady, predictable growth is being stretched sorely.
• Firms are now focusing at a more granular level ' practice group, office, areas of law, timekeeper ' and these subsets of the firm do not uniformly reflect the firm-wide growth rate and ratios.
• Most importantly, this approach provides neither a benchmark against which to measure actual performance nor a means to understand and analyze monthly variances.
• Under the typical approach, if in April we collect only $8 million instead of a budgeted $8.5 million, where can we look for an answer to why? Is the variance driven by less-than-expected production in April or a prior month? Is it realization-driven, either for billing or collection? Or is it only a matter of timing cutoffs, to be made up later? Answers to these questions are not readily available, even at a firm-wide level, because we have assumed these linkages away.

Modeling How Work Value Translates to Cash

Fortunately, there is a more reliable way to conduct revenue modeling. By using historic details of billing and collections that show how work value actually translates into cash, firms can create more accurate, multidimensional (granular) forecasts ' by timekeeper, practice group, office, etc. This improved approach, outlined below, takes firm-specific patterns into account and allows for work value budgets to be adjusted accordingly:

Access historic billing and collections information. Most time and billing systems today are SQL-based, making it relatively straightforward to construct and run queries. Using 3 years' worth of history:

• Select each month's work value and each subsequent month's billings related to that month's work.
• In exactly the same fashion, select each month's billings and each subsequent month's collections related to those specific billings.
• Add dimensional query filtering as needed: by timekeeper, practice group, office, etc.

Convert. For each individual month, and for the most recent 2- and 3-year averages by calendar month (eg, 3-year January average), convert these dollar amounts into percentages; eg, of the $10 million work value in January, what percentage was billed in January, in February, etc.? Most firms do have a fairly predictable seasonality to their billings and collections: January billings are low, December collections are high, etc.

Note: Redwood's Data Warehouse product performs these calculations automatically. Running similar calculations manually is also straightforward, however, once the SQL queries are properly set up.

Select appropriate patterns. Will we be budgeting based on last year's pattern, a 2-year average or a 3-year average? What is the most relevant comparison? For example, if 2 years ago our practice expanded substantially through lateral additions, or the entire firm grew as the result of a major merger, then last year's numbers might be the only indicative set.

Apply the patterns. Use the patterns to develop the monthly billings in dollars of each month's work effort and the monthly collections in dollars of each month's billings. The resulting collections budget is thus built from the bottom up. In the spreadsheet illustration below, for example, the $10.7 million of Total Collections budgeted for February is based on $4.0 million from January work, $1.1 million from February work and $5.6 million from accounts receivable open at the beginning of the year. When February's actual results are available, we can compare these balances to those in actual as described below.

[IMGCAP(1)]

Repeat the above steps to calculate the analogous patterns for liquidating year-end WIP and A/R to cash.

Analyzing Variances

In addition to more accurately predicting cash flow at a more granular level, the approach just described greatly ameliorates the problem of not being able to perform proper variance analysis. Returning to our example of collecting only $8 million in April instead of a budgeted $8.5 million, we now have the tools necessary to understand what's going on. As long as we continue to pull the data described above, we can answer the following important questions:

• How has each recent month's production compared to budget?
• What is our cumulative total billed (or collected) from each month and how does that compare?
  If our production is strong and the cumulative percentages are low, we can deduce a timing issue (although we still need to act). If our production is off and percentages are higher than budgeted, we have a real problem that can only be overcome in all likelihood by stressing greater production and better realization.
• Alternatively, we can look for offsetting expense savings. Such changes are hard to effect mid-year; but by identifying the problem earlier than we ever could before, we will at least have given our firm the chance to make those changes.

Paramount among the many analytic challenges facing Law firm CFOs and their financial staff is accurately forecasting cash. Relying on the law of large numbers, most firms assume that prior averages will hold, so they use history-based, firm-wide performance ratios to obtain cash flow projections.

Simplistic Forecasting

Accurate information is essential for effective forecasting, however, and historically many firms have taken a simplistic approach to getting at this information:

• They determine work value by timekeeper (number of hours x standard rate).
• Then they apply realizations (amount billed against standard, amount collected against amount billed) to determine the collectable work value.
• At this point, a number of variations on essentially the same approach are used. Prominent among these is to spread this collectable value across the previous year's collection pattern to derive budgeted collections.

For example if we budget $100 million of work value at standard, historically bill 96% of standard, and historically collect 98% of billings, then we will budget collections of ($100 million x .96 x .98) = $94 million. If in January of this year we obtained 10% of our annual collections, we will budget collections for January of next year at $9.4 million.

In a growing firm, this simplistic approach has an inherent weakness: over-budgeting revenue. The problem is that not all of this year's growing production will be collected this year. It takes several months to collect on work, so that some of the increased revenue will be booked but not available as cash until the following year. Moreover, firm growth usually occurs gradually over a year, so that the firm shouldn't expect to see much of the total expected annual revenue increment in the earlier months of the year being budgeted. It's not these patterns themselves that cause the problem, but that firms aren't accounting for them in their shortcut forecasting process.

Many firms utilize a historic haircut to right this wrong. For example, if the firm is growing at a 10% annual growth rate, they will reduce the annual budgeted collections by up to 10%. This approach previously worked just fine for firms, but is now problematic for three reasons:

• In today's consolidating and competitive environment, the formerly reasonable assumption of steady, predictable growth is being stretched sorely.
• Firms are now focusing at a more granular level ' practice group, office, areas of law, timekeeper ' and these subsets of the firm do not uniformly reflect the firm-wide growth rate and ratios.
• Most importantly, this approach provides neither a benchmark against which to measure actual performance nor a means to understand and analyze monthly variances.
• Under the typical approach, if in April we collect only $8 million instead of a budgeted $8.5 million, where can we look for an answer to why? Is the variance driven by less-than-expected production in April or a prior month? Is it realization-driven, either for billing or collection? Or is it only a matter of timing cutoffs, to be made up later? Answers to these questions are not readily available, even at a firm-wide level, because we have assumed these linkages away.

Modeling How Work Value Translates to Cash

Fortunately, there is a more reliable way to conduct revenue modeling. By using historic details of billing and collections that show how work value actually translates into cash, firms can create more accurate, multidimensional (granular) forecasts ' by timekeeper, practice group, office, etc. This improved approach, outlined below, takes firm-specific patterns into account and allows for work value budgets to be adjusted accordingly:

Access historic billing and collections information. Most time and billing systems today are SQL-based, making it relatively straightforward to construct and run queries. Using 3 years' worth of history:

• Select each month's work value and each subsequent month's billings related to that month's work.
• In exactly the same fashion, select each month's billings and each subsequent month's collections related to those specific billings.
• Add dimensional query filtering as needed: by timekeeper, practice group, office, etc.

Convert. For each individual month, and for the most recent 2- and 3-year averages by calendar month (eg, 3-year January average), convert these dollar amounts into percentages; eg, of the $10 million work value in January, what percentage was billed in January, in February, etc.? Most firms do have a fairly predictable seasonality to their billings and collections: January billings are low, December collections are high, etc.

Note: Redwood's Data Warehouse product performs these calculations automatically. Running similar calculations manually is also straightforward, however, once the SQL queries are properly set up.

Select appropriate patterns. Will we be budgeting based on last year's pattern, a 2-year average or a 3-year average? What is the most relevant comparison? For example, if 2 years ago our practice expanded substantially through lateral additions, or the entire firm grew as the result of a major merger, then last year's numbers might be the only indicative set.

Apply the patterns. Use the patterns to develop the monthly billings in dollars of each month's work effort and the monthly collections in dollars of each month's billings. The resulting collections budget is thus built from the bottom up. In the spreadsheet illustration below, for example, the $10.7 million of Total Collections budgeted for February is based on $4.0 million from January work, $1.1 million from February work and $5.6 million from accounts receivable open at the beginning of the year. When February's actual results are available, we can compare these balances to those in actual as described below.

[IMGCAP(1)]

Repeat the above steps to calculate the analogous patterns for liquidating year-end WIP and A/R to cash.

Analyzing Variances

In addition to more accurately predicting cash flow at a more granular level, the approach just described greatly ameliorates the problem of not being able to perform proper variance analysis. Returning to our example of collecting only $8 million in April instead of a budgeted $8.5 million, we now have the tools necessary to understand what's going on. As long as we continue to pull the data described above, we can answer the following important questions:

• How has each recent month's production compared to budget?
• What is our cumulative total billed (or collected) from each month and how does that compare?
  If our production is strong and the cumulative percentages are low, we can deduce a timing issue (although we still need to act). If our production is off and percentages are higher than budgeted, we have a real problem that can only be overcome in all likelihood by stressing greater production and better realization.
• Alternatively, we can look for offsetting expense savings. Such changes are hard to effect mid-year; but by identifying the problem earlier than we ever could before, we will at least have given our firm the chance to make those changes.