Campuskhana is a lesser known food tech startup operating in the food capital of the city, Indore. While giants like Foodpanda are making losses manifolds their revenue, here is a startup which has been profitable and has been running successfully for the last 15 months generating a total revenue of about INR 2 million while fulfilling more than 6000 orders and getting around 2000 satisfied customers onboard. Let’s hear the story straight from one of the co-founders, Amit Ranjan as to how CampusKhana created a profitable Food Tech startup while others were running into losses?
I once wanted to send a birthday cake to my sister in a medical college with considerable student strength. To my surprise, there was no online player which was doing this. There were 4000 students on campus. On an average, there would be 11 birthdays a day and yet no player to cater to this audience. On closer analysis,I found out that this campus was still living in the stone age as far as online world was concerned. There were no online food, grocery, fruit, bakery delivery services. As I discussed this issue with other people, I realized that this issue was prevalent in so many institutes. Most of the premier institutes are located in tier-2 cities or outskirts while most of the online aggregators are located in metropolitan cities. While in cities like Bangalore, you are spoilt for choices, many educational institutes have none. This real life problem became the basis of my pilot. I teamed up with Monica, a postgraduate student from IIM Indore. When we started in December(2015), most of the food tech startups were bleeding profusely and hence our whole idea was to make this pilot profitable. There were problems galore; for starters, people believed that if Zomatos or Foodpandas of the world were not operating in these institutes, it must be a loss making avenue to venture out. Here are the specific details of how we managed to remain profitable-
The Chosen One: selecting the institute
We had to zero down on an institute where we could pilot this idea. We created a list of 20 probable institutes, mostly engineering and management, based on following factor-
- Are there any other major competitors in the market? No.
- Are there atleast 1000 students staying on campus? Yes
- Proximity with other educational institutes to expand.
- Is the institute accessible easily? Yes
- Purchasing power of the target audience (higher the better)
While the first 4 factors are easy to gauze, there in no established data to compare purchasing power of one institute over the other. We assumed that there would be some correlation between the fee charged by an institute and the purchasing power of the students. More an institute charges, richer is our target audience. The last factor eliminated all the engineering colleges and we were left with management schools. Among all the IIMs,IIM Indore has one of the highest batch strength owing to its integrated program on top of the regular post graduate program and executive program. It also had few other institutes nearby.
Solving the logistical mess: Analog to Digital
With all pomp and show, on 6th December(2015) we went live. We were not ready for the overwhelming response that we got. We realised that we had to either find a way to regulate the demand side (orders that we receive) or the supply side (deliveries that we make). We first tried to regulate the demand side. We started scheduled delivery where we would accumulate orders for a given period and then process the entire batch together. This had two problems-
(i) Food delivery is a time critical business. No one would place an order which would be delivered in a slot which is 2 hours from now. While such models might work for grocery delivery businesses like Big Basket, food delivery is a different ball-game altogether. Our orders dropped by 63% when we started scheduled delivery.
(ii) Food orders have huge surge during lunch and dinner time. Hence if I have a slot ,say, from 8:30 PM to 9:30 PM, most of my orders will be placed in this slot. Hence, resources required for this slot could be as much as 300% of the off-peak period.
The idea of scheduled delivery failed miserably. This is when we focussed on the supply side. We realised that our delivery fulfillment was analog; orders could originate from anywhere in 5 KM radius and would have to be delivered anywhere in the the 193 acre campus. To simplify this logistical mess, we had to turn digital. We needed to restrict the restaurants that we cater to and the delivery points. While initially we had onboarded any and every restaurant, we now identified 6 of the most popular restaurants. Luckily for us, Top 5 of these formed two distinct clusters due to close proximity. We had to let go of a restaurant for time being. We then got few other restaurants in these clusters onboard. Similarly, we had 5 distinct delivery clusters. So, from an infinite possibility, we now had only 10 possible paths. Turning from analog to digital was a great move for us in solving the logistical mess.
Being operationally profitable in all transactions
Our ultimate goal was to be profitable. We took it a step forward to try and be operationally profitable for all the orders that we deliver. This was a herculean task as the audience that we were catering to was very price sensitive. While we wanted to be operationally profitable, we could not do that by levying the operational cost as delivery charge for orders below a certain ticket size like Swiggy does. This certainly placed a ceiling to the delivery charge we could levy. We started with free delivery above Rs 150, with order below Rs 150 charged a nominal fee of Rs 15. Certainly things don’t add up here. If I am fulfilling a Rs 160 order, restaurant would be paying me somewhere around Rs 30, but my delivery cost would not be less than Rs 45. This is where the simplified delivery network came into picture. Now, since every order placed in a particular time interval from one restaurant cluster to one delivery cluster could be clubbed, we could easily club orders. We have set up smart algorithm to optimise the order clubbing so that it does not impact the user experience. On an average, our delivery agents carry 2 orders in one trip, hence reducing the overall operational cost by 50%.
Just to check the efficiency of Swiggy’s delivery network, I recently placed two orders from the same restaurant to the same address 5 minutes apart. To my surprise both these orders were assigned to two separate delivery guys even though the only difference in those two orders was of the mode of payment, surprising but ‘none of my business’.
Optimum use of delivery Network
Food order demands are not uniform. There two major spikes, one during the dinner time and the other during the lunch time. These spikes could be as high as 300% of the normal orders. In addition to it, there would be spikes during weekends and holidays. These spikes again could be as high as 200% of the weekday orders. So for a weekend dinner time, you have to be ready for 600% surge. How do staff yourselves in such situation. If I hire delivery agents for the peak time, more than 95% of the time my delivery network would be under-utilized. At the same time if I am not prepared for these 5% of the time, most of my users would have bad experience and would never come back and would further dampen my already damped normal orders. So the only option left out is to prepare for the peak orders but to also ensure that the delivery network is optimally utilised. We achieved this in following 2 ways-
(a)Employing extra work force during the peak time:
Somewhat like Uber surge pricing, we introduced incentives so that extra delivery agents could join us during the peak time.
(b)Diversifying the offering
To further optimise the delivery network, we diversified into grocery, fruit and cake delivery. Hence, during the off peak time, we could use our workforce for these.
No discount, Cashback only (that too when necessary)
Food Delivery as a business is a low margin game. It is very unlikely that you would be profitable inspite of the discount. One might write it off as customer acquisition cost which would be outweighed by the Life Time Value (LTV). But the problem here is that loyal customer is a myth. Entire market is very price sensitive. The customers acquired by discounts vanish as the discounts vanish. Cashback also is a loss making proposition. But there is silver lining. Firstly, cashback increases the stickiness of the user. If you offer cashback, users have incentive to comeback and transact.Secondly, effective cash burn is lesser than what it appears. If I provide 10% cashback on a 500 order, I provide user with Rs 50 Campuskhana credits. To redeem these, user will have to place another order, let’s suppose of same ticket size. So the effective cashburn would only be 5% in this case.
While we rarely ran instant discount campaigns, we encouraged partner restaurants to offer discounts on our portal. We would provide them insights as to time of the day or day of the week when their share of order was not as expected. There were restaurants which used this data to run customized discount campaigns on our portal.
Incentivizing Online Payment
In our context, Cash on Delivery had following problems
- Cash exchange between customer and delivery agent can be time consuming
- Managing cash flow is resource and time consuming process.
- First time COD orders need to be authenticated
We were catering to a limited audience, but at times we would receive order from places that were beyond our area of coverage. Most of these were COD orders.With online payment, people would not put fake address just to ensure that order goes through.
We aggressively incentivized online payment. As we are opposed to deep discounting, we provided CK credits for online orders. Hence reducing the cash transactions and increasing the stickiness of the user at the same time.
Let the data speak
We have a relatively smaller data set, over 600 users and approximately 1400 guest accounts to be precise. Over a period of time, we have tried to read into these data. We have data set of people who have only ordered dinner from us and not lunch. We have data set of users who have only ordered from one particular restaurant multiple times. We have data set of people who always eat alone and those who always eat in big groups. These data sets help us customize our offerings and promotions. As we grow, we would have even more data to play with.
Existence as USP
The greatest advantage that CampusKhana had was that our existence was our USP. Let me explain this. There was no competitor who was doing anything that we were doing. We were the pioneer of online food delivery, grocery delivery, fruit delivery and cake delivery in our space. We could attract customers just by our mere existence. To get the returning customer, we just had to ensure that we provided a great user experience. Unlike most other food tech startups, we could both keep the cake and eat it. Pie was all ours. Hence there was never a price war and hence never an urgency to provide discounts.
While we started as a food delivery platform for IIM Indore, we subsequently diversified our offering by including fruit, grocery,cake etc. We also introduced ‘Bring This’, using which users could hire a delivery agent to get anything fetched from anywhere. We also expanded to other campuses like Indore Institute of Science & Technology, Indore Institute of Law, Medi-caps & Silicon city. While we expanded and diversified at the same time, we made sure that we did not open all our war fronts at the same time. We would start something, trump it, let it stabilize before trying out something else. While we walked slow, we ensured we never lost our ground.When we look back today,we realize we have covered good distance over time.
In food delivery business, what we make is just the tip of the iceberg. Most of the profit remains with the restaurants and rightly so as they are the creators. To crack this, we started our own branded mealbox. While we did not own kitchen, we procured food from one of the restaurants and branded it as our offering. We had to put quality checks in place and provided standard packaging on the top. But this ensured that we could keep 45%-50% of the revenue
While I have painted a rosy picture of food-tech startup, it has its own problems. Unlike any electronic good, margin of error in food industry is very low. Food has to be delivered to the user in a very small time frame after which it is useless both for the user and vendor. Product is not standardized. Hence utmost care has to be taken to screen the restaurants that get onboard. One or the other item could be out of stock in the restaurant and real time inventory management is difficult. Managing fleet of delivery agents on the road is a herculean task. Till now we have managed to tide over all these hurdles. But this is just the beginning. While we have managed to remain profitable on a smaller scale, maintaining profitability when we scale up is yet to be seen. We are up for new set of challenges in the upcoming months. We are screening institutes in other cities where we can replicate this success model. We also have college specific user behavior pattern and are exploring options to leverage that.