99% of Restaurants FAIL – How Petpooja built a software to save them? : Business Case Study
Executive Summary
Parthiv's story has four mechanics. First, the pivot: his 2012-13 corporate meal service (400-500 orders a day for banks like HDFC, Rs 80,000 in daily orders at 10-15% margins) broke on working capital - corporates paid in 45-60 days while restaurants wanted 30 - and on the missing tech layer at restaurants, which meant four phone calls per order and 1,600 calls a day: 'we were basically running a call center.' Second, the research: instead of surveys, he watched small restaurants work - a bill taking 90 seconds because legacy software choked on un-archived databases - and foresaw the FoodPanda/Zomato/Swiggy wave needing a connected, cloud-based restaurant OS. Third, pricing: from free (with Idea dongles handed out so merchants could get online) through Rs 1,000, 2,000, 5,000, 10,000 a year - each raise added more customers, until Rs 15,000 triggered abuse and a rollback to Rs 10,000 plus taxes. Fourth, the moat: an open, Android-style marketplace of 300-plus integrations where Petpooja is the merchant's first call even when Zomato is down, and where integration partners spent 2017-2020 pitching Petpooja to their own customers.
Key Takeaways
A working-capital trap killed a growing service business
The corporate meal service scaled nicely - 400-500 orders a day, HDFC as anchor client, 60% of clients from financial institutions - but corporates paid in 45-60 days while restaurants demanded payment in 30, so every incremental order consumed more cash. Combined with Parthiv's Gujarati instinct that 'cash should come into your hand first,' the math made growth self-defeating: the more business he won, the more capital he needed to survive it.
Four calls per order means you run a call center, not a company
With no software at the restaurant end, every order meant at least four phone calls - confirm the menu, remind them to cook, check packing, chase the delivery boy - so 400 orders generated over 1,600 calls a day. Scaling would have meant scaling a call center linearly. That missing tech layer, not demand, was the real bottleneck, and it pointed directly at the product Petpooja became.
Observe, don't survey: the 90-second bill
Instead of questionnaires, Parthiv's team stood in small restaurants and watched: a dine-in bill took 90 seconds because each menu item took 15-20 seconds to select on legacy desktop software bloated by an un-archived database. High-volume stalls like a dabeli wala making 2,000 bills a day couldn't tolerate that at all, which is why they used mechanical bill-makers. Merchants never report these pains - to them it's 'the standard process' - which is exactly why observation beats asking.
Sell to the vada pav wala, not to McDonald's
Parthiv picked SMB/MSME restaurants deliberately and calls it product-market-culture-cofounder fit: a suit-and-boot enterprise salesman cannot sell to a dabeli wala, and the stall owner cannot sit across from a VP of IT. Define your target group and stay in it for the three-to-five years the product needs to mature; selling enterprise one day and SMB the next doesn't work in the early life of a company. The upside: nobody else was pitching software to these merchants at all, so Petpooja owned their mind space.
At zero, everything looks good
Petpooja's price discovery climbed a staircase: free first, then Rs 1,000, 2,000, 5,000 and 10,000 a year - and at every raise, onboarding increased, because a price made customers take the product seriously ('it's available at the right price, take it'). Then Rs 15,000 hit a wall: furious calls, abuse on email, sales tanking, customers threatening to rip the software out. The rollback to Rs 10,000 plus taxes taught him the leverage equation Ganesh quotes from an IIM-A founder's talk: leverage equals the value you provide minus your price - and you can only raise price while no cheaper alternative delivers the same value.
Android, not Apple: let competitors onto your platform
Instead of locking merchants in, Petpooja became an open restaurant OS: a marketplace of 300-plus integrations including rivals' CRMs, loyalty tools and website builders. The payoff was trust and distribution - when Zomato goes down, the merchant's first call is Petpooja, whose support tools already show the outage, and between 2017 and 2020 the integration partners themselves were pitching Petpooja to prospective customers. The trade: you give up some lock-in and per-app revenue, you gain the merchant's mind space and a word-of-mouth flywheel.
The next gaps: inventory photos, payroll, robots
Parthiv's forward map is still observation-driven: only 25-30% of even Petpooja's clients (and maybe 10% of the wider market) do inventory properly, so invoice-photo digitization now puts food-cost and P&L reports at a merchant's fingertips; payroll ran on Excel and pen-and-paper even where fingerprint devices existed, so Petpooja built integrated hardware-plus-software; and robotics, which once seemed fictional, now runs delivery robots for eight-hour stretches abroad - he expects automation in every facet of the food industry, starting with simple loops like a chutney-and-sambhar robot in Udipi restaurants.
The Conversation Breakdown
From a pohe shop to the corporate meal layer
⏱ Untimed topicGanesh opens with discovering Petpooja on the billing screen of a Pune pohe shop and frames his thesis: India is software-deprived at the grassroots, and restaurants fail not from bad food but from inefficiency. Parthiv traces his path: three years running an F&B outlet during his engineering days in the late 90s (every job except chef), then the 2012-13 corporate food service with co-founder Apurva - registering as the single vendor for banks and financial institutions, aggregating 300 Ahmedabad restaurants into budget-based menus, and solving the trainers' daily complaint of identical food. HDFC tested them for three days straight; after that, 'we have HDFC' sold every other institution.
Why they walked away: working capital and 1,600 calls a day
⏱ Untimed topicTwo structural problems ended the service business. First, cash: corporates settled in 45-60 days, restaurants demanded 30, so growth mechanically increased working-capital needs - and chasing payments offended the founders' Gujarati cash-first instinct. Second, tech: with nothing digital at the restaurant end, each order took at least four phone calls, meaning over 1,600 calls a day at peak - 'it felt like we were running a call center.' Choosing where to build, they picked the restaurant side after walking the SMB market and finding dabeli and vada pav sellers using no tools at all.
Market research by observation, and the cloud conclusion
⏱ Untimed topicParthiv's method: don't arrive with a questionnaire - merchants will only name today's pressing problem - go and watch. They watched a bill take 90 seconds on legacy software slowed by an un-archived database, saw stalls using mechanical bill-makers to avoid software entirely, and reasoned forward from TechCrunch and GrubHub: when the FoodPandas of the world arrive, merchants drowning in tablet orders will need a tech layer, that layer must connect to the outside world, and therefore it must be cloud. Corporate-sales lessons carry over: find the department that owns the pain, cold-reach the person, and meet face to face, because a prospect's confusion is visible on his face and invisible on a call.
Adoption hacks and the three tailwinds of 2014-17
⏱ Untimed topicEarly adoption was brute force: the software was free, and when merchants said they had no internet, Petpooja handed out Idea dongles. Then three forces converged: competition rose and margins thinned, forcing owners to run multiple outlets; owners stopped sitting at the billing desk, making theft-by-unbilled-orders a real fear (Mast Kalandar pioneered 'no bill means the meal is free' to force billing discipline); and Jio plus the Zomato/Swiggy/FoodPanda wave made internet ordering mainstream. The biller juggling a dine-in customer while a Zomato order silently cancelled became the canonical pain - Petpooja's integration auto-accepted the order and printed the KOT with no human in the loop.
Pricing discovery: the staircase, the abuse, the rollback
⏱ Untimed topicStarting free, Petpooja raised prices stepwise - Rs 1,000, 2,000, 5,000, 10,000 a year - and paradoxically signed more customers at each step, because a price signalled a real product. At Rs 15,000 the story broke: angry calls and emails ('I'll rip your software out and throw it away'), threats to bill manually at new outlets, and tanking sales forced a rollback to Rs 10,000 plus taxes. Ganesh adds the IIM-A framing: leverage equals value minus price - Zoom saves Rs 50,000 of travel for Rs 50 a month - and the ceiling on your price is set by the cheapest alternative delivering the same value. Post-COVID, while two heavily funded competitors dragged prices down, Petpooja still raised 18% two years running.
The OS play: 300 integrations and the first-call trust layer
⏱ Untimed topicPetpooja chose Android's philosophy over Apple's: an open marketplace where the merchant gets the best tool for each job even if Petpooja didn't build it - 300-plus integrations including competitors' CRMs and loyalty programs. The discipline behind it is support: they built internal tools so their teams know when Zomato itself is down, making Petpooja the merchant's single first call instead of a blame loop between vendors. The flywheel: from 2017-2020 integration partners pitched Petpooja to their own prospects. The product stack grew in workflow order - Captain app for waiter ordering, table reservations with RSVP prompts, Kitchen Display Systems routing KOTs by cuisine section, loyalty, website builder, waiter-calling hardware, payroll hardware - now 25-28% of revenue versus 75% from the POS core.
Gaps everywhere: hotels, robots, and the Gujarati risk gene
⏱ Untimed topicAsked for the next five years, Parthiv points at undigitized basics before fancy robotics: invoice-photo inventory, payroll automation, and only 10% of the market measuring anything - 'if you're not doing inventory correctly, you don't know whether you're making money.' He and Ganesh swap pain stories about ITC and Marriott check-out queues and Indigo's payment counters as proof that every industry has visible, solvable gaps - hotel ERPs, auto check-in, luggage robots. He closes on Gujarati business culture: children grow up watching fathers and uncles run businesses, struggle is expected and cyclical, and failure carries no taunt - even the failed uncles say 'go ahead, why not' - which makes risk-taking feel default. His parents opposed his startup; his uncles backed it.
Core Frameworks & Ideas
Observation over interrogation
Customers can only name today's pressing pain; they have normalized the chronic ones. Watch the workflow instead - time the bill, count the calls, notice the unused fingerprint device. Every observed inefficiency, solved in the order the customer meets it (reservation, ordering, kitchen, billing, loyalty), becomes your next product.
Leverage = value - price
Price is capped by the customer's savings: Zoom at Rs 50 against Rs 50,000 of avoided travel can raise prices almost freely; a product with a free or cheaper same-value alternative cannot. Quantify your value in the customer's money, check the cheapest substitute, and the gap between them is your pricing headroom.
The open-OS trust trade
Choosing marketplace openness over lock-in costs per-app revenue and hands customers to competitors occasionally - but it makes you the customer's single first call, turns integration partners into your unpaid sales force, and compounds as a word-of-mouth flywheel. Sequence matters: earn the core workflow first, then open the gates.
Product-market-culture-cofounder fit
Match the target group to who you actually are: the SMB stall owner and the enterprise VP of IT require different founders, different salespeople and different products. Pick one TG, stay for the 3-5 years the product needs to mature, and only then expand - enterprise one day and SMB the next kills young companies.
Actionable Next Steps
- ✓Founders: this week, go watch three customers use your product or process end-to-end and time the steps - Parthiv's entire roadmap started with a 90-second bill he saw, not a survey he ran.
- ✓Before your next price increase, write down the customer's savings in rupees and the price of the cheapest same-value alternative; the gap is your headroom, and outside it you get the Rs 15,000 treatment.
- ✓Restaurant and retail owners: if you cannot see today's sales, bills and inventory without calling your outlet, treat that as a leakage emergency - start with billing discipline ('no bill, meal free') and inventory tracking, even if it's just invoice photos.
- ✓Anyone entering a legacy industry: map the customer's full journey first interaction to last (reservation to loyalty, booking to check-out) and pick the most evident inefficiency as your entry product, as Petpooja did with billing - the gaps in hotels, airlines and hospitals are sitting in plain sight.
Notable Quotes
“If you are not doing your inventory correctly, you cannot even have your efficiency right - if that is not correct, then you don't know whether you are making money or not. (translated from Hindi)”
“It felt like we were basically running a call center. (translated from Hindi)”
“Between 2017 and 2020, it was not only us pitching Petpooja to prospective customers - all of these partners were pitching Petpooja to prospective customers. (translated from Hindi)”
“At zero, everything looks good. (translated from Hindi)”
The One Thing to Remember
The deepest lesson: go and watch, don't ask. Parthiv's product map - cloud POS, KOT automation, the Captain app, payroll hardware, invoice-photo inventory - came from standing in restaurants and noticing what owners had stopped noticing (a 90-second bill, a fingerprint device nobody used, four calls per order). Merchants don't know their own inefficiencies; they call them 'the standard process.' Observation finds the pain a questionnaire never will, and each observed inefficiency, solved in workflow order, becomes the next value-added product.