Par Technology management has been coy about software revenue line items and gross margin, but we did the digging
In March 2021, Par Technology's CEO commented that non-software service revenue carried a 30% gross margin
Can you give me a rough split between the gross margin and software versus gross margin on field and warranty services, just to get a sense of the difference in profitability?
Savneet Singh•PAR Technology Corporation•CEO, President & Director
Yes. So we historically haven't -- in MD&A, we give a lot more specific here, but we haven't broken it out. I would say that our software businesses will be very much like other software products to sell to the same ACV levels that we do. We should be 70% to 80% gross margin in time with our product. We aren't there today, primarily because as we've talked about in the past, the business has run incredibly inefficiently, and we've sort of had excess spend.
But we see a really nice trajectory in gross margin growth for the next couple of years.And I would expect us to be in that range over time. I don't expect this to not be in that range as we -- I don't see us being any different than any other software product that sells at the price point that we sell to. And so that's already happening. And again, the new products we built, the new products that we've acquired have all come in that sort of 70% to 80% gross margin range. We should be no different.
On the non-software side of our services revenue, you're in the 30s of gross margin, and that is somewhat volume dependent. So as we grow Brink, that business line sells more hardware, so more services are affiliate with that. And that -- and I think it's very traditional that sort of hardware business that package software. Installation implementation is not a business that we make a ton of money in on the gross margin level, but warranty and exchange, those field services is actually where we make nice margins.
And so as we grow, I think that business line will -- that part of the margin line will stay relatively flat. As growth slows, you'll see also increased margin there, because you're doing less implementations where the gross margin is relatively low.
Hopefully the CEO misspoke
Assuming Mr. Singh is half correct and that the non-software service revenue carries a 15% gross margin, one can work backward and mathematically calculate LTM software gross margins. Doing so, suggests Mr. Singh is either way off on non-software service revenue, or software gross margins are currently very low, equating to 42% on a trailing basis, down 25% from last year.
Math doesn't lie. Either non-service revenue gross margin is lower than 15% historically or software gross margins are currently VERY low and in a downward trend
That being said, despite stating software margins would carry long-term gross profit margins of 70-80%, $PAR management has been transparent that software margins are lower today and that those margins are aspirational. It would appear there remains much to aspire to.
Consistent with the idea that software margins are VERY low, software revenue as a percentage of service revenue has increased from 14% in 4Q17 to 44% in 1Q21, but service margins are only 4-5% better.
If you '0-out' margins for the non-software service business, it implies LTM software margins of 60%. Any way you cut the math, this is a low margin software business today before the addition of Punchh, with Brink and Restaurant Magic seemingly carrying combined gross margins of +/- 40-50%.
Take Me to the Brink
Up until the recent Punchh acquisition, Brink has been the focus of $PAR and its growth story despite weak revenue growth
Monthly Brink software revenue per average site is down 9% y/y potentially signaling pricing pressure
Pricing pressure in software would be consistent with service gross margins staying flat despite mix of software tripling
Using the company's own disclosure, growth has stalled dramatically in recent quarters
Has Brink been giving pricing concessions in recent quarters to win or retain business?
Quarterly revenue and average sites are utilized to calculate monthly Brink software revenue per site
Assuming all 3,327 Brink sites in backlog are completed during 2021 at an average revenue per site of $157 per month (last reported amount) and 6% annual churn (lower than the recent 9-10% churn), Brink ARR growth in 2021 will be roughly 9%
10% +/- growth with 50% margins and considerable competition suggests a single-digit multiple, 7-8x $25 million revenue
Brink is likely only worth $200 million
Can Restaurant Magic Pull a Rabbit Out of Its Hat?
Don't hold your breath
$PAR paid around $45 million for Restaurant Magic in December 2019. Based on disclosure in transcripts and filings, management has suggested Restaurant Magic revenue since deal closure has taken a flattish path
1Q20 revenue = $2.0 million
2Q20 revenue = $2.1 million
3Q20 revenue = $2.2 million
4Q20 revenue = $2.1 million
1Q21 revenue = $2.25 million
Like Brink, Restaurant Magic has seen software revenue per average site drop in recent quarters. Given the implied low gross margins across all software 7-8x multiple on a no growth, low margin business seems generous.
But lets be even more generous and say RM is worth $100 million, 2x what $PAR paid despite no growth
If Punchh is Worth $500 million....
$PAR Software Business Seems to Be Worth
500 + 200 + 100 = 800 million
Current pro forma enterprise value is more than double that at $1.7 billion
$123 million cash at Q1 end
$288 million PF debt
25.8 million PF shares
Backing out $250 million value for the Gov't biz = $1.4 million PF value
Backing $200 million value out for hardware biz = $1.2 billion
$400 million overvalued / 25.8 million shares = $15 downside (low $40s)
This assumes Punchh value remains steady despite the destruction in software; using Punchh's revenue data from $PAR, top-line growth decelerated from 100% in 2018 to 50% in 2019 to just 19% in 2020; clearly CV19 made an impact but the amount is uncertain
At $32mm ARR, PAR paid 15x for Punchh; a 5x contraction in multiples alone would impair the deal by $150 million
Restaurant Magic & Total Software
What's in That ARR Anyway
Given the aforementioned low software margins, one has to wonder how much "related revenue" is included in $PAR ARR
We define Annualized Recurring Revenue (ARR) as our annualized revenue from Subscription as a Service (SaaS) and related revenue of our Brink POS line of business. We calculate ARR by annualizing the monthly recurring revenue, or MRR, for the last month of each reporting period.
PAR's credit agreement definition of ARR raises even more questions
“Annual Recurring Revenue” shall mean, with respect to a Test Period, the sum
(x) all revenues derived from software as a service and related recurring support services revenue for the last fiscal month in such Test Period multiplied by twelve (12),
(y) the monthly average amount of revenues derived referral fees for software as a service and related recurring support services for the three fiscal months in such Test Period multiplied by twelve (12) and
(z) the monthly average amount of revenues derived from merchant transaction fees for the three fiscal months in such Test Period multiplied by twelve (12), each as recognized in accordance with GAAP, but excluding the impact of any purchase accounting or other adjustment arising out of the consummation of the Closing Date Acquisition and historical or future acquisitions.
Notes and Links
There are a number of instance where $PAR management changes the reported recurring software revenue. As an example on the 2Q20 call,
2Q19 Brink recurring revenue was $4 million, whereas on the 2Q19 call they say it was
. There are a number of similar instances going through transcripts and filings.
Management seems to mix recurring software revenue with recurring service revenue; the mix could help explain low "software" margins
The LTM as of Q3 we were at $10.5 million of software revenue; about $3 million of recurring service revenue, which is a very sticky recurring service revenue.
Most of the reported backlog seems to be
RPO / deferred revenue from service contracts
related to hardware and software installations
, where many of the Brink and RM revenue figures were reported