gday. 1. ASR (LVT's equivalent of ARR) is based on signed...

gday.

1. ASR (LVT's equivalent of ARR) is based on signed subscription agreements. If we owned a company with ASR of $40 million, we could fire all the employees except those actually operating the servers and receive an annual revenue of $40 million. Overtime this would gradually decline to nothing due to churn , which is why churn is such an important valuation metric. Of course, negative churn, would see the revenue increase without doing anything.

2. However there is a lag between a subscription agreement being signed and receiving subscription revenue:
(i) Software sales are typically heavily bunched towards the end of a quarter. Salesmen scurry to make quarterly targets. Customers take advantage of end quarter deals.
(ii) The software then has to be installed and the vendor will always provide a training and installation period, before charging starts
(iii) Quarterly or monthly SAAS is billed in arrears. The payment terms typically range from 30 days to 60 days. You pay your bill 30-60 days after the end of the quarter.
In March, LVT signed approximately $2.6 million new subscription agreements. We guestimate they will be installed between April and May. Assuming they are a quarterly contract , the first three months invoice will be approx August 1st, payable on approx October 1st.
(iv) To make matters more complicated, some contracts will be an annual contract, paid up front 60 days after the software is installed and training complete. These contracts will often include an incentive (because of the better cash flow) and can be the first 14 months for the price of 12 etc etc. That is why you see the bump above merely taking the growth in ARR , lagging it 6 months , and dividing by four. Using a regression of growth in revenue vs growth in ARR, we estimate that approx 30% of LVT's subscription contracts are annual contracts. We bounced this number off management....they did not disagree.....but honestly it is only an estimate and we observe that annual contracts seem to be gradually declining (which we happen to like)

From all this you can build a complicated model to forecast revenue where the inputs are:
- Actual revenue from one year ago
- 4 quarters of ASR growth lagged 6 months
- An input assuming a mix of annual and quarterly contracts (try starting with a 30/70 mix)
- Churn rate assumption. Since growth of ASR captures the positive side of negative churn, you need to avoid double counting but there will be some lagged affect between shrinking customers (immediate decline in revenue) and growing customers (lagged increase in revenue). The economy is growing so not many customers are shrinking and LVT seems a very sticky product so the number turns out to be inconsequential for LVT but its a standard part of any model (and highlights the importance of monitoring churn).
- Add something for non-subscription revenues. LVT has a professional services revenue which is not subscription based.
- We also have a line item for the N3 revenue, which is approximately 850,000 per quarter and is included in LVT's ASR (so we take it out of one line and add it back in another).

Until Wizdom came along, our model accurately forecast revenues within $75,000 of actual declared revenue looking six months forward. The model improves as it learns about contract mix and payment terms, For Sept and December, actual revenue was within $10,000 of forecast revenue. Wizdom disrupted the model but the model will get a better handle on that in this quarter and next.

The beauty of SAAS is that we can see revenues six months out based on ASR announcements now.