Survey Says…

As you may know, near the beginning of the KitchenPC design phase, I conducted a survey to try to get a grasp of who my users would be and how they behave.  I’ll conclude up front that while I think this attempt was an overall failure, as I don’t believe it really led to smarter development decisions, the data I got out of it are still worth using as kindling for the cognitive fire.  To everyone who took the survey, I do greatly appreciate your time and honest feedback.

So, let’s go over the answers, shall we?

In total, I got 276 full responses to the survey, and another 15 partial responses.  While this is far from my initial target of 1,000 responses (which apparently requires incentives I just can’t produce at this time,) I think the data show a fair sample of the “techie” family types that are liable to try out a product such as KitchenPC.  So, what types of cooks answered the survey?  Only 8 would identify themselves as “gourmet” cooks.  57 rate themselves as an accomplished cook, and 154 (a little over half) are comfortable with the “home cook” label.  33 are aspiring, and 31 say they need help.  While these are simply labels that one assigns to themselves with no common denotation, the summary I can extract from this question is that I can be fairly comfortable trying to target the average home cook, which is nice because that’s what I consider myself.  One can also raise the valid point that the people who took my survey are not necessarily who will end up using the product.  Maybe I’ll attract gourmet cooks from all around the world on day one, and end up with full page adverts in “Gourmet Cooks Monthly.”  However, the strategies I used for finding survey respondents will be similar to those I employ to find beta users, so I predict a similar demographic for initial users.

Do you use recipes?

Do you cook using recipes?  This was a rather important question.  I mean, if you don’t use recipes you probably would get almost no use out of KitchenPC.  Luckily, I already know there’s about 55 million people in the U.S. that use recipe websites on the Internet, and culinary startups are huge these days.  Still, only around 5% admit to always using recipes when cooking, and 25% saying they often do.  The majority, with 51% of the responses, say they’ll use a recipe as a starting point but make their own changes.  This trend is a bit disappointing to me, as KitchenPC favors the cook who wants to plan out meals and compile a definitive list of what they need at the store.  If they’re tossing recipes out the window when culinary inspiration strikes, I might be a bit out of luck.  Still, I can hope that such modifications of recipes involve changing amounts, tossing in a few extra spices lying around, or leaving things out that don’t sound good.  In geek terms, these changes happen at “run-time” and will be independent of disciplined meal planning.  Even if this is making a wrong assumption about this user type, providing a service to the nearly 30% of cooks who usually use recipes is still a viable business.

The next question is a biggy, as it seeks to validate pretty much the entire principal behind why the site was created.  Do you usually plan meals in advance?  I’m actually somewhat pleased with the results, however I think the question was very poorly designed.  57 people (about 20%) said yes, which is awesome!  I would love to graph a correlation here to find out who these people are and how I find them.  Do they have families?  Are they experienced cooks?  These are my peeps!  However, 95 people (32%) said no.  The majority answer, at 129 people (44%) was “Sometimes”.  This is somewhat aggravating as I wish I had broken down this question to get a bit more detail about this.  Still, from these results I can ascertain that there’s a good chunk of people out there that would use features like a meal calendar, being able to compile accurate shopping lists from multiple recipes at once, etc.  The industry backs this conclusion with the number of online meal planning sites popping up on the Internet like weeds (or in the occasional simian form.)  Obviously, entrepreneurs think there’s a market out there for this, and I’m one of them; I just feel no one has really got it right yet.  I’m a bit comforted by the fact that KitchenPC will provide features for non-planners that I think would be useful, such as the ability to figure out what to cook based on an available set of ingredients.  I think this sort of feature can actually encourage those “sometimes” people to plan in advance a bit more.  Since multiple meals can be “modeled” quickly and easily, users can type in the main items in their fridge they’re trying to use up, and KitchenPC can suggest a few meals for the next few days.  Someone who answered “No” to this question now all of a sudden finds themselves with a meal plan that saves them money, targets their likes, and encourages them to try something they haven’t cooked before; ha, I’ve made a meal planner out of them!

What's preventing you from planning meals in advance?

The next question only popped up for those people who answered “No” to the above question.  It simply asks, “What’s preventing you from planning meals in advance?”  The obvious trend in the answers (as well as those who typed in their own response) is that people just don’t think about it.  After reading these answers, I get the feeling that these people live alone and cook for only themselves.  I’ve made this conclusion purely due to the first person singular viewpoint in the responses (“when *I* eat” or “when *I* shop.”)  Hopefully I can play around with these data more in Excel and try to confirm this conclusion.  Needless to say, people who don’t plan meals in advance simply don’t want to; they figure out what they’re in the mood for right now, or by what looks good at the store, or just blatant procrastination.  There was a lot of references to “feel like” and “mood” in the responses.  Apparently, people aren’t military generals driven by a strict schedule.  “0700 Hours – WAFFLES, PEOPLE!”

It’s obvious KitchenPC needs to provide something for non-meal planners.

Opinions on meal planning

So how about the people who do meal plan (at least sometimes)?  38% of them say they do it to save money and optimally use ingredients.  Bingo!  These people are gonna dig this site.  I’m all about saving money and telling you how to use that leftover 13oz of pork you have.  51% say they do it to save time by consolidating their shopping list for the week.  Double bingo!  I do exactly that, and I do that ridiculously well.  25% say they don’t have time to think up meals on the spot.  This contrasts with the non-planners who somehow don’t have a problem with thinking up meals on the spot.  Like I said, I think it’s all about finding out what types of people plan meals.  The easy conclusion is that if you do plan meals, you’re gonna love KitchenPC.  The 20% of people who do plan meals in advance easily provides a large enough demographic to build a profitable business around.  Especially since the pool is so large to begin with.  20% could easily be millions of people, if marketed right.  I’m excited by this prospect, enough to justify my original business vision of building the best meal planning tool on the web.

How do you decide what to cook?

The next question is a very simple one; “How do you decide what to cook?”  The majority (65%) say they try to use ingredients that they have on hand.  This is great news, as most of the innovation around KitchenPC is “reverse meal planning”, i.e. finding meals that will use what you have on hand.  Nothing on the web that I’ve seen can do this well, or involving multiple recipes at once.  26% of people say they choose a cuisine first, depending on their mood.  I think these users will still get some benefits from KitchenPC, using modeling based around recipe tags, which focus on the major cuisines.  136 people (46%) say they just make whatever sounds good at the time.  I believe these people will also like the meal planner as it’s simply just a recipe suggestion engine that works with whatever data you give it.

Do you use a shopping list?

“Do you bring a list to the grocery store?”  If you do, you’re like 54% of my survey respondents.  A dismal 6% say no (I found this surprisingly low), and 37% sometimes do.  I think these numbers definitely support the need for an intelligent shopping list creation tool, one that’s easy to use, flexible, and integrates with mobile devices.  However, how do people create shopping lists?  Only 34% say they keep a running list through-out the week, with 42% saying they jot down their list right before going to the store.  Now, peoples’ habits are almost impossible to break but I can’t help but think better shopping list tools might tend to entice some people to store their running shopping list online, especially those doing a lot of meal planning.  Only 7% fell for my humorous answer, “I wander the aisles aimlessly asking myself what I need.”

The survey also asked users if they eat frozen or pre-packaged meals.  The idea behind this was judging the potential for designing some features around this concept.  I’m going to go out on a limb here and say that even people who eat pre-packaged meals are not the ones planning up front.  “Hmm, yes I shall eat this Red Baron pizza on Wednesday the 17th.”  In case you’re curious, only around 3% eat these meals almost every night (probably while watching some sort of reality TV show,) with 22% admitting to this habit a couple times a week, and the majority (30%) once or twice a month.  I really don’t see much value from this information.

The next few questions involve online shopping.  One of the revenue models being considered post beta is integration with online grocery stores.  Shopping list generation is done so accurately, it’s viable to implement a system that can transfer your shopping list to an online grocer such as Safeway.com or Amazon Fresh, similar to purchasing flight tickets through Expedia.  Needless to say this will not be a feature offered any time in the near future, however I’m trying to get the data to judge its feasibility (as it would be a massive business venture requiring many inter-company partnerships.)  To let you guys in on a little secret, I actually have a working prototype of this feature and a mocked up online grocery site called MikeMart.  I’ll blog about it more post-beta.

93% of people have purchased stuff online.  This is no surprise.  I also asked if people purchase groceries online.  Only 33% said yes, with an overwhelming 65% saying never.  Out of the people who do purchase groceries online, only 6 (not percent, 6 actual people) do so once a week or more.  11% do so once or twice a month, 42% do this every few months, and 33% have done this once and only once.  40 of these people have used Amazon Fresh, and 33 have used Safeway.

Why don't you buy groceries online?

Why don’t people buy groceries online?  I think the plethora of dot-bomb ruins over the past decade speak for themselves.  This is a near impossible business model, and people are just not going for it.  While it is growing slowly, the statistics I dug up showed that only around 1% of Americans purchase groceries online.  Still, companies such as Amazon Fresh are doing well, as well as Safeway who is also making a considerable investment into this market.  62 people said they don’t buy groceries online because it’s too expensive (e.g. delivery fees), and 115 said they want to select the items themselves (in other words, they don’t trust what’s gonna show up on the truck.)  37 people chose to write in their own response, and their answers led me to believe these people will not be easily convinced to change their ways.

I think the idea of trying to support KitchenPC from online grocery shopping revenue is far fetched, however it’s still a feature I would love to pursue at some point if I get enough users to convince Safeway or Amazon Fresh to sit down and talk with me.

What factors do you consider while grocery shopping?

When asked what factors people consider while grocery shopping, 71% said price, 67% included value, and 79% checked the box for quality.  Very few people (18%) cared about brand names.  Almost half mentioned they look for organic food, care about farming ethics, etc.  I’m actually quite interested in this concept, and plan to write a more protracted summary of these thoughts in a future blog post.

Another revenue model to explore is coupons (hey, it’s a multi-billion dollar business!)  Only 4% of people admitted to always using coupons, but there was a very nice bell curve here.  10% said often, 51% said sometimes, 13% said at least once, and only 16% said never.  I think very few people are going to turn down a coupon if it just magically appears on their printed shopping list, and it’s as easy as just scanning in at the grocery store.  For this reason, I think selling coupons through KitchenPC is a viable option to consider as a revenue model, and probably doable in a way that won’t annoy all my users.  The fact that I completely normalize ingredients in my database makes this implementation very straight-forward.

How many people do you cook for?

Here’s a question that purely exists to try to correlate results with other answers.  On average, how many people do you cook for?  I’m very curious if cooking for more people makes you more likely to meal plan, however those are data I’ll try to tease out once I get into the marketing stage.  15% only cook for themselves.  42% cook for themselves and one other person (young, married couples I guess?) and 31% cook for 3 of 4 people.  Only 7% cook for over 4 people.

I also asked people if they use club rewards cards.  21% do for a single grocery store, 53% have these cards for multiple stores, and 15% claim they have a rewards card for every possible store in their area.  Only 7% don’t use these cards.  This question was once again targeted around grocery store integration, determining if it would be necessary to transmit member information along with orders, and displaying member pricing as well.  Obviously, this would be a must.

Would you share what you cook?

The last question was geared around privacy, as the original prototype had a lot of social networking features that unfortunately will not make it into the public beta.  Never-the-less, it’s interesting information.  “Would you be willing to share what recipes you cook with others?”  This was a rather poorly worded question, but what it means is do you care about people seeing your calendar.  53% say “Yes, with anyone and everyone.”  20% say they’ll share this information with their friends and friends of friends, 9% saying friends only, and only 4% fear big brother watching and wouldn’t be comfortable letting anyone know what goes on in their kitchen.  It’s pretty easy to see that privacy won’t really be a huge concern, however the post-beta version of KitchenPC with social features will definitely include some basic privacy controls.

Luckily, 178 people left their email addresses for me, which will be great as I’ll send them invitations to the upcoming private beta.

Overall, I think I got some valuable data from this survey, but it’s a far cry from any real ammunition for basing an entire business plan on.  I’ve found my mind is pretty attached to my existing business plan and tends to look for the conclusions I want to see from the numbers.  As I’ve said earlier, I think customer driven development is a very valuable tool for certain types of businesses. If you’re spending millions on a prototype, and have a burn rate of $500k a month, you’d better prove beyond a shadow of a doubt that you have a customer base, show a demand for your services up front, and have the raw data to back it up.  For a site like mine, I think a more reasonable goal is to just get the damn thing online.  It’ll hopefully be useful to a few people, and those people will show their support by coming back.  The real question is agility.  When you do realize you didn’t understand your customer, how fast can you fix it?  Can I figure out what types of people are using my product and devise a marketing strategy to attract more of them?  These are things you have to do with a real, live product and not some half-assed survey.  At least that’s my theory, and I’m sticking with it.

Again, thanks to all of you who took the survey!  It was fun to try out and I’m glad I did it.  I think being an entrepreneur is about trying a bunch of stuff and seeing what sticks, but being honest when something doesn’t work out the way you want.  Who knows, maybe another reader out there will be able to use my data to influence a business idea of their own.