Sales forecasting may be the most inexact science of all business disciplines. Or it used to be. Because, if you lead a high-tech sales force of 70 or more reps, Aviso president Steve D’Angelo has some good news for you. His artificial intelligence (AI) software can give you much more accurate sales forecasts – and with far fewer headaches for sales operations staff and managers.
Furthermore, Aviso offers much more than just pinpoint quarterly predictions. The tool’s pipeline analytics let execs and managers drill down to see weekly pipeline changes, flag inconsistencies in rep versus data-driven forecasts, and generally make execs and managers much smarter about what is happening in sales and why. That means they can act on issues and changes before they become problems.
Sales forces with 40 or 50 reps probably do not need the sophistication of Aviso, D’Angelo acknowledges – even if they would benefit somewhat from its improved accuracy. “But, above 70, the problem gets too complex,” he says.
Indeed it does – because a good sales forecast needs to incorporate plentiful data provided by reps, the judgment of seasoned managers at several levels, and, ideally, whatever sophisticated data analysis is possible.
The way forecasts are done now, even in giant and very sophisticated corporations, is generally something like this. Reps enter data in their CRM systems, and this data is extracted from CRM and entered into spreadsheets by sales operations. Managers (sometimes at several levels) use their experience to handicap prospects, and these adjusted predictions are rolled up to the next levels by sales operations staff.
One problem is that errors can accumulate in the roll-up, as handicapping formulae are not always passed on to the next level. Second, the whole process takes a lot of effort – sometimes keeping sales operations staff working over weekends for a big forecast on Monday. Finally, it’s all very anecdotal and depends on the unaided wisdom of reps and managers.
Aviso takes a different approach. Every 15 minutes, the software takes a snapshot of every opportunity in the sales pipeline. All this data is entered in a time-series database. AI tools use behavior and the results of past deals to generate a forecast. D’Angelo says the beginning-of-quarter Aviso forecast will be 90-95 percent accurate by quarter’s end. And Aviso can predict sales well beyond a quarter.
The Aviso approach takes the messy, manual, and problematic approach to sales forecasting and makes it automatic and much smarter. Yet it is not a black box that execs must accept without understanding. “We show every step,” D’Angelo notes.
Managers just click on a region, a territory, or a product and see the deals that support the overall forecast. They can drill down to individual reps and see who is expected to hit their numbers. And they can compare rep forecasts with AI-driven predictions.
Managers can also drill down to individual deals and see how highly AI scores their likelihood of closing. Red flags and green flags show either problems requiring attention or clear sailing. “They have a lot less insight into the forecasts when they are using spreadsheets,” D’Angelo observes.
All these capabilities are based on correlations learned from the actual experience of the company using Aviso. For example, when an opportunity stays at a certain pipeline stage longer than normal for winners, it is flagged as problematic, based on what happened to similar opportunities in the past. That is also a warning to reps.
Managers get it all – the top-line numbers and the detail, the comparison of rep versus Aviso forecasts – on a dashboard updated constantly.
Aviso has been impressing some large, very sophisticated and fast-growing companies with its capabilities.
For example, 14-year-old Splunk is now doing $1 billion a year of business making software that searches, monitors, and analyzes machine-generated big data. After very rapid growth, a fifth of the firm’s sales staff was busy managing forecasts in spreadsheets, yet these spreadsheets did not give visibility into deals or show changes in the pipeline.
Aviso freed up those number-crunching reps for real sales efforts and stopped wasting everyone’s time on defending forecasts. Accurate and automated forecasts let reps and managers focus on strategizing and closing deals instead. And Aviso gave visibility into the real-time health of each opportunity in the pipeline. Spunk CEO Doug Merritt says he is “thrilled by the insights Aviso has brought to me and my team.”
Or take Apttus, which has been growing fast with its quote-to-cash software and is now worth about $1 billion. Hyper growth and a constantly-changing organizational structure were complicating sales forecasting. In addition, the company wanted a forecasting tool that worked with multiple CRM systems and improved accuracy and speed to make Apttus attractive in an IPO.
The firm got all that and more from Aviso. “Aviso provides accountability, visibility, and predictability in our sales forecasting process,” summarizes VP of business operations Ankur Ahlowalia.
Another client, nine-year-old Aryaka, has been growing its connectivity business globally and last year landed in Gartner’s Magic Quadrant for wide-area-network optimization. Things had not always looked so smooth for the company. When Shawn Farshchi became president and CEO of Aryaka in 2015, he could not get an unbiased and reliable analysis of the sales pipeline. Sales forecasts were rarely accurate and were, invariably, out of date. Farshchi considered one major forecasting tool, but found it “bare bones; I knew it wouldn’t fulfill our needs.”
The new CEO learned that Aviso could give him and his new team insights into pipeline quality at both corporate and deal levels – right at their fingertips. “Our company is moving fast and quick decisions need to be made,” Farshchi stresses. “When the board needs an update, I need to quickly get answers. Aviso enables this.” And speed came with accuracy: the first-quarter Aviso forecast came within 7 percent of actual results.
Aviso’s data-driven approach does rely on CRM pipeline data. It is stronger the more reps enter all the appropriate data in their CRM system. But D’Angelo says Aviso can work with “dirty data” from some reps – filling in entries and calling managers’ attention to reps who need to improve their CRM use.
Aviso is hosted in Amazon Cloud, so it requires no IT infrastructure. Implementation begins with a few hours to extract two years of history on opportunities won and lost. This data is entered in the time-series database, and machine learning techniques are applied. Data scientists tune and tweak forecasting algorithms to improve predictive accuracy. The whole process takes about 30 days for a major high-tech firm.
The process of accurately forecasting past sales often persuades Aviso clients of the tool’s value. “Wow, you nailed it, one user said,” D’Angelo remembers.
This kind of accuracy is not just a product of mathematical wizardry. It also reflects Aviso’s decision to focus initially on one industry vertical: high tech. “Every industry is different: finance, health care, manufacturing, retail,” the Aviso exec says. “We wanted to focus on one sector first and really master it. In coming years we will probably do other industries.”
The 30-year software veteran believes he and his company are in a pretty good and unique place. Other software firms are just beginning to get into sales forecasting in a serious way and usually emphasize pipeline management more than actual forecasts. Since they’re not the same but are interdependent, forecasting using AI can positively impact both.