Forgot to top off your dish with fresh cream? Get it in 10 minutes, they say. Ever wondered how?
Cooking is certainly a skill a lot of us have picked up during the pandemic. We started cooking more and were seeking fresh groceries more than ever to cook delicious food. And, if you can’t relate then, here’s some proof.
As you can see from the ‘recipe’ and ‘grocery’ Google Trends charts for the last 5 years in India, we started searching for recipes like crazy during the lockdown-induced 1st wave of Covid. To cook recipes we were trying for the first time, we were keen to add the best ingredients. Thereby, resulting in an uptick for grocery as a search term too as shown below. These charts clearly show that the onset of the pandemic caused a paradigm shift in consumer behaviour.
Two years back in 2020, we were ordering groceries online from BigBasket, Grofers and Amazon Fresh due to the fear of stepping out. These apps were not able to catch up with the rising demand for online grocery orders. I remember not being able to get flour for making rotis from one of the apps above. As a last resort, I had to make do with ready-to-cook rotis. Even if the products were available, you wouldn’t get delivery slots even on the next day. It would probably get delivered a week later in some cases. This frustration somewhere created a need for a certain cross-section of the population to get groceries as quickly as possible. This is what ushered in drastic shifts in how we shop for groceries and brought Quick Commerce to prominence like never before.
The rise of Quick Commerce
To understand the impact of shifting consumer behaviour, let’s see how the approach to getting groceries is changing for some people.
The frustration that I mentioned before has led to the following behavioural changes:
Saving money (value) → Saving time (convenience)
Monthly bulk orders → Weekly micro orders
Planned grocery shopping → Unplanned/impromptu ordering
Order upon exhaustion of commodities → Top-up orders in advance to avoid waiting for items to come back in stock
Note: The above behaviour doesn’t apply to all categories of products but major ones.
All of the above have given birth to the Q-commerce or quick commerce market. Q-commerce penetration within the online consumables market is ~7% and is expected to grow to 12-13% by 2025. Consumers include Gen Z, ambitious millennials, Gen X-active buyers and Gen X- Passive buyers. Quick commerce was estimated at USD $~0.3 Bn in 2021 and is expected to grow 10-15x in the next 5 years, to become $5 billion by 2025.[1]
What is Quick commerce though? Quick commerce is defined as the delivery of consumables within 45 minutes with a nominal delivery charge.[1]
The above shifts in grocery shopping behaviour are on the rise and we can confirm from the below chart that there's a steep rise.
Instamart by Swiggy started delivering groceries in 15-30 minutes in 2021 and I was just amazed by how fast they would deliver anything. And today, we have players like Zepto, BlinkIt (formerly Grofers), Ola Dash and Dunzo getting up to speed in this market. Time to address the big elephant in the room - the dark stores that are enabling this industry.
What are Dark Stores?
The word 'dark' doesn't have a negative connotation. They are called so because customers can't walk in and buy anything. They are exclusively used only to fulfil online orders. Only the employees (i.e. packers and order handlers) are present and delivery guys pick up orders from there. This is not a new concept and is being followed globally by retailers like Walmart and new players like Gorillas.
You won't see big billboards or branding to help you find them. They are like micro-fulfilment centres and not huge like warehouses. For example, places like abandoned cinemas and marriage halls are used as dark stores in India.[2] As you don’t need a huge space, it is possible to have these stores very close to the customer they are serving. And, I suspect that only a set of delivery folks are allocated to a certain dark store.
Imagine delivery apps not using dark stores for a moment. You hire shoppers that go to nearby stores (similar to Swiggy Genie & normal Dunzo pickup services). They will have to navigate the stores to find each item. There may be some items that are not available. Also, normal customers will be present too and would cause delays due to crowding. On top of that, some more time will be wasted at the billing counter too. It would probably take 5-10 minutes to just get the order and leave.
How are some orders getting delivered in under 10 minutes?
10-minute delivery makes us think about the plight of delivery folks who we think would be resorting to rash driving to meet the order delivery deadline. However, with dark stores, every customer is not more than 3 km away. Some may be just 500 metres or 1.7 km away that saves a lot of time and allows you to pick up more orders. I mentioned 1.7 km because I see my Instamart order being picked up every time from the same location which is approximately that distance. These stores are strategically placed in areas where the order density is high. High order density means the number of orders in that area is very high compared to areas with low order density.
But, with a deluge of orders how do you ensure that everything is packed for every incoming order without any delay?
The secret to this success is optimising the lead time it takes for an order to be made ready for delivery. Here’s how the lead time is calculated -The clock starts ticking as soon as the order is received once you checkout and it stops when all the items are picked up from the shelves, packed in a bag and made ready for the delivery guy to pick up.
The above process takes just 58 seconds on average for Zepto, as claimed by Aadit Palicha, co-founder and CEO of Zepto.[3] This process would take much longer as I mentioned before but is now happening within a minute.
Currently, the median delivery time for Zepto is around 8 minutes 47 seconds.[3] This is not the average time and this is where a lot of people confuse the central measures of tendency (mean(average), median and mode). Median is like the middle of a range. For example, there are 5 orders and you sort their completed delivery times in ascending order as shown below.
Delivery Time for 5 orders (in ascending order):
5 minutes 10 seconds
7 minutes 39 seconds
8 minutes 47 seconds
10 minutes 21 seconds
16 minutes 3 seconds
And, you could see the delivery time in the middle of this range is 8 minutes 47 seconds, which is the median. So, it doesn’t mean that all orders are delivered within 9 minutes. It means that if 8 minutes 47 seconds is the median delivery time for like 100,000 orders a day then there are almost 50% orders delivered in less than that time!
How is this even possible? You can find some articles that show you the inside of these dark stores. There are no automated mechanical devices packing everything. The employees of the dark store only pick up the orders from shelves and pack them. If we think about what would be driving such high efficiency, then the following areas come up:
Dark store floor map (i.e. where are the shelves, how big is the area of the store, where will be the pickup counter, where will the order monitoring system be, etc.)
Segregation of product categories across the store shelves in such a manner that it takes minimum time to pack them
Order and inventory management
I am sure they are using a lot of data points to optimise each of the above areas. So, let’s go back to 58 seconds - the time taken to dispatch an order after a user has placed the order. To achieve this lead time, the employees of the dark store have to be super quick to get the orders ready for the delivery pickup. Not a single motion can go to waste by the store employees while getting the order packed. There is a study to measure and optimize how you can complete a given task in the most optimized manner. It is called time and motion study.
To visualise this optimization process through complex data points is a humongous task and my guess will hardly touch the surface. So, here’s an easy way to help you perceive this process through a tennis court choreography scene from the movie ‘The Founder’:
As you can see, you don’t need to make the store employees work harder or faster. The dark store floor map is designed in such a manner that the packers take minimal time to make the order ready. This optimized store layout ensures that not a single motion goes to waste. That’s how the quick-service restaurant McDonald’s became a huge success. Permutations and combinations of each step were tested to ensure no waste of a single motion to get a fresh burger ready to serve.
Similarly, the same process refinement is relevant for dark stores. Without any customers, you can ensure the lead time is as low as possible so the delivery folks don’t have to rush.
What does the future look like?
Other players are also stepping up their game. Ola Dash is adding 500 dark stores across 20 cities, BlinkIt has added 200 dark stores. There’s no doubt that we will see more dark stores coming up.
One thing to notice is that these stores are not very big. They are small and house around 2000 products across categories. What happens when the dark store is not able to catch up with demand? It can and does happen.
Here’s an anecdote of my own experience: A few days back, I was trying to order a Whiskey Sour cocktail mixer. I just checked whether it is available on Zepto. It wasn’t but they had other 2 flavours in stock which I didn’t want. Then, I visited Instamart and I found it and placed the order. Yes, these apps are suitable for such micro-orders with nominal or no delivery fee. Ordering from the official site of the brand lets you buy a minimum 8-pack order of 250 ml mixers, which is inconvenient. But, on these apps, I can just order one 250 ml mixer with no delivery fee for a party of two to enjoy after a long day. Similarly, for different products the availability is not always the same. This makes you switch between the apps.
Through my above experience, I realise that getting inventory management right for such micro orders is still something that needs to be mastered by these apps. I know this is a challenge as the above items are part of long-tail categories (i.e. Soda & Mixers) as opposed to categories like Dairy, Fruits & Vegetables, etc. This is called product range efficiency which is one of many factors that will decide which quick commerce players will be able to achieve long-term profitability.
Hope you found this helpful!
References: