Minimum Viable Product · Product testing

Validating a same-day delivery app?

Jason Yu

May 18th, 2015

Hi everyone,

I'm in the early stages of working on Instacart for ethnic grocery delivery. As a Taiwanese-American living in San Francisco, I find it hard to find basic ingredients for Chinese cooking and, having spoken to other Chinese, Korean, Indian, Russian friends, etc., am realizing that many others feel the same way. And for my Caucasian friends who enjoy ethnic cooking, I sense that going to Chinatown or an Asian supermarket (i.e. 99 Ranch) can be a confusing and intimidating experience (i.e. labels in other languages, less-than-helpful staff, generally out of the way compared to Safeway/Whole Foods, etc.)

At this point, I'm interested in quickly testing something. What is the best way to size the market and potential for on-demand or same-day delivery type apps in the bootstrapping stage? 

Chris Hartman Experienced Product Development Consultant

May 20th, 2015

Hi Jason,

You mentioned 3 main problem areas:
1. Inconvenience
2. Confusion
3. Context

How much customer discovery have you done to validate that those are actually the major problems, and if so, who has them? I think getting some quantifiable data on that would help steer you in a direction where you may more easily determine what sort of prototype could validate a solution.

I'm also not sure if you are targeting people who regularly buy ethnic food, or casual foodie types who just want to test the waters with something new. The former sounds like a big market to crack, but one with staying power, while the latter seems finicky and likely to jump on whatever the latest hot app will provide.

So far, it looks like you are moving towards an on demand delivery service, which seems loosely related to the problem of Inconvenience. That is, "it is a pain for me to drive across town when I need the ingredients now."

In my personal experience, I feel the main issue is that nobody offers a complete selection across all ethnic categories. For example, in my area of West Los Angeles, there are a number of Japanese markets and several South Asian/Indian markets. However, if I want to buy Chinese products I need to go downtown, Korean to Koreatown, Filipino to Glendale etc...  Even if there was a Ranch 99 nearby, it just doesn't solve the problem of selection across all types of food. We buy a lot of ethnic food. If I could get it all delivered without driving all over town that would be a huge win for me even if it wasn't instant. But that is me, and I'm not you and you're not everyone else, so I'd start with learning a bit more about who you are trying to serve first.

Try talking to a broad set of users in different demographics and ask them what their biggest problem is in relation to ethnic food. Maybe target a sample of people who have shopped at an ethnic grocery store anytime in the past 12 months. Learn about their habits. How do they cook, what do they need most often, how often do they shop and where. Is is perishable? Do they stock up? etc.. Are there any common threads between the types of people who buy Korean groceries vs. Chinese? Does that even matter? Are there any specific "ethnic" categories that have the largest need? 

You may also want to validate the assumption that people who shop at ethnic groceries are older and that the problem of context is valid. If your primary target market are those who cook that type of food often, understanding the labels may not be a problem worth solving. Don't forget about the potential for foreigners living in the US, or recent immigrants. Consider foreign college students. I see the Japanese, Chinese and Korean markets filled with younger people here in LA and many of them are foreign students. 







Michael Brill Technology startup exec focused on AI-driven products

May 19th, 2015

Jason, 

OK, if you stop thinking about your value as delivery then it opens up a bunch of things.

Market size aside, an Online Chinatown with recipe and descriptive content layered over products aggregated from a couple dozen vendors sounds frigging cool. For some reason, if I shift my mental model from Ranch 88 and delivery to Chinatown and culinary exploration, it's a qualitatively different thing. A little bit like www.goodeggs.com, but obviously different products and more educational content. You have a lot of arbitrage possibilities to get started because pricing is quite opaque. As a consumer, I love the idea of someone grabbing a duck from a window, combining it with gai lan, garlic and black bean paste I need for a recipe I saw on your site, oh... and you just happen to throw in a couple of custard buns in the bag to surprise the customer. 

You can take this a bunch of different directions... add in more prepared foods so I can get instant gratification (take on a big chunk of Chinese delivery), create personalities for your vendors for authenticity, and so on. Tons of experiential elements. Start a Blue Apron-type service for the Chinese diaspora or people who visited San Francisco but live in some crappy place where they can't get anything not made by Kraft.

Instacart is about saving time (certainly not money - depending on what you buy, you will pay between 50% and 100% surcharge vs a smart shopper in Safeway). That's not you. You are about experience first and saving money second. You're not helping people avoid Chinatown, you're bringing it to them online in a fun, low-stress way. 

Michael Brill Technology startup exec focused on AI-driven products

May 19th, 2015

What is "hand rubbed" beef? Is there a machine rubbed beef? And what, exactly, does rubbing beef do? ;-)

Freeman Fan Entrepreneur

May 18th, 2015

That's an interesting idea Jason. I've been thinking about building something like that, but in the prepared-food space instead of groceries.

A couple of market sizing methods that I found very useful were taking out some Google Adwords or Facebook ads in your target geographies, and then gauge interest by looking at the click-through rate. Even better, build a simple landing page with a sign-up-for-beta form, and gauge the conversion rate.

My worry is that (1) there might not be enough demand and (2) people who shop at ethic supermarkets are less technology savvy. My parents who always shop at Chinese supermarkets find Uber and Instacart difficult to understand, while I (also a first generation immigrant) shop at Ralph's and such way more than at Chinese supermarkets.

Michael Brill Technology startup exec focused on AI-driven products

May 18th, 2015

Seems like metro logistics is/is going to be a hyper-competitive space with everything available from everywhere. You're also likely to run into the situation where people need a jar of weird stuff every now and then vs. weekly staples. OTOH, you may have a niche in delivering crab paste and durian... but say goodbye to your dating life.

Not to be officious, but maybe something closer to Freeman's idea... prepared foods or an "ethnic" home meal assembly business... a long tail Hello Fresh / Blue Apron?

You can run tests and convince yourself of a market opportunity with a BS metric ('ooh, look, I have 7 people who want this therefore it's business), but if your core value proposition is moving something from point A to point B, that's a kill zone.



Jason Yu

May 19th, 2015

Hi everyone,

Thanks for the valuable feedback - really appreciate it!

Karl Schulmeisters CTO ClearRoadmap

May 19th, 2015

Michael - there is a new startup here in Paris doing kindof what you just suggested https://eatpopchef.com/

They guarantee delivery of a fresh cooked lunch in 15 minutes. They change their meal every day- and now are offering two dishes to choose from. To deal with logistics they are using the ubiquitous Scooter Delivery - IE the Pizza Delivery model

Essentially they (and this business that Jason is looking at) is a Delivery Pizza business. So to be successful you need to look at who is successful in the Pizza Delivery business and how.:

Domino's is probably the biggest. And how do they do it?


  1. Limited menu
  2. fast delivery
  3. limited geographic delivery
  4. scale by franchise.
  5. High margin/low COGS on products offered


The challenge I see for an "ethnic grocery delivery" is that you violate #1- which in turn has impact on #2 (filling the order takes more time the more items you have)

and #3 is challenging because by definition "ethnic grocery" is going to have a smaller market in any given area, than something else. So to get your volume to a sustainable level you either have to

  • expand your offerings - which violates #1
  • expand your geographic reach - which violate #3 and therefore #2

#5 also seems like a challenge - because a lot of the more esoteric items are esoteric precisely because their margins don't justify a large grocery store like Safeway  in giving the item any floor space.

Seems to me a very challenging business to be in

Jason Yu

May 19th, 2015

Sorry if this is a repeat - looks like my last comment was accidentally deleted...but the thing is you can still allow users to get whole recipes delivered, including "regular groceries" you can find elsewhere.

Grocery prices in a Chinatown for "regular" things like onions, lettuce, chicken, for example, are in general much cheaper than what you'd find at any grocery store. Even things you wouldn't expect like avocados are available and cheaper in Chinatown.

So it is also an arbitrage play of sorts, where even if you were to charge a higher premium for delivery (i.e. 20-25% + fee + tip vs. Instacart's 15% + delivery fee + tip) to make up for greater inefficiencies in the delivery model, the final cost to you is likely still less than or equal to what it'd cost to buy a similar basket of goods at a big box store (without delivery).

Karl Schulmeisters CTO ClearRoadmap

May 18th, 2015

Well one thing to look at is all of the 'same day grocery delivery' companies that went bust in the Dot Com bubble burst.    Many died because they simply could not generate enough revenues.  

And at the same time you have a lot of large grocery chains offering this sort of capability,  And an company like QFC or Safeway can easily add the contents you are going to offer.  So how are you going to differentiate?

Michael Brill Technology startup exec focused on AI-driven products

May 19th, 2015

Karl, this model (home meal assembly) has really taken off in the US with companies like Blue Apron at > $100M annual revenue. One thing you see here is that services are increasingly bundling up, say, 4 meals in one shipment which materially lowers the delivery costs and logistics complexity (now you have a few days to deliver instead of 15 minutes).

But that's a different offering from Jason's original idea which more of an Instacart model which really isn't so dependent on a small number of SKUs. His point is that there are plenty of people who shop at Asian markets for most of their groceries (markets like 88 Ranch are bigger than a typical French supermarché) and that they want convenience just like people who shop at Safeway. Seems logical that there would be, but it also seems that it's quite small and that their demo is older and less likely to avail themselves of a delivery service (especially with a higher proportion of fresh ingredients where human judgment comes into play for selection). *Probably* younger people (Asian heritage or not) just need a jar of fermented squid eyes that Safeway doesn't sell... the problem then is that you're in the business of delivering condiments and lack the volume of products necessary to justify this model. Even if you could figure something out, you'll be crushed by a horizontal metro logistics service company... maybe even Uber. ;-)

There is plenty of room to innovate in food with this as the problem statement: http://theoatmeal.com/comics/cook_home