Day 155: The Tinkerdash reviews are in, I didn’t quite expect this

“The more I live, the more I learn. The more I learn, the more I realise, the less I know.” – Michel Legrand

Sadly, the one-month Tinkerdash blogger review was now over. The good news, however, was that four of the reviews had been published and we were delighted by how much we were able to help people and learn from the experience.

As Claire, one of the reviewers put it, the one-month Tinkerdash pilot had taught us a lot of things that we never expected it too.

Yes, it was expected that Tinkerdash would help people save time and money. It was not expected, however, that it would also help families spend more time together or help people understand how food affects their body.

From the feedback, it was also suggested that people would be willing to pay in the region of £20-£30 for the “full” service, which included the adding of the ingredients to the customer’s online supermarket basket. Useful to know but yet to be validated by a bunch of people signing up and paying this figure of course.

And not all of the feedback was positive, for example, a lot was learnt about the scalability of the concept (or lack it), as discussed in a previous post. It also became evident that the onboarding process needed some work, amongst other aspects of the service such as planning over longer periods than one week.

On whole, it proved to be a highly beneficial experience and one I’d recommend as a means of getting early feedback from users. Check out this previous post for details on how I set it up.

To read what four of the reviewers had to say in full, visit their personal blogs at the following links:

Day 153: I decided not to wing it (for once), here’s why and what I did instead

“It seems that bad advice that’s fun will always be better known than good advice that’s dull no matter how useless that fun advice is.” – Scott Berkun

As brought to light in a recent BBC documentary titled “Clean Eating’s Dirty Secrets”, there’s an unprecedented amount of nutritional advice available online nowadays. And worryingly, not all of it is good for us nor are all of the people giving the advice qualified to do so. For some, large social media followings seem to have replaced actual qualifications as the measure of whether somebody’s advice is to be trusted.

From an ethical standpoint, let alone business, this was not something I wanted Tinkerdash to fall guilty of. Any services provided should be helpful to people and not potentially harmful.

So, I decided to get in touch with an expert. Using, I sourced the contact details of five registered dietitians in the UK and contacted them, explaining that I’d like to work together to try to offer people a top-notch service.

(Check out this article if you’re unsure of the difference between a dietitian, nutritionist and a nutritional therapist. I opted for a registered dietitian given that their profession is regulated and hence they would each be trained to a set standard.)

From the two enthusiastic responses I received, it seemed that a consultative approach would be most suitable. The registered dietitian could provide guidance for creating meal plans to cater for a range of specific dietary needs and assist with ensuring the recipes and plans were balanced in terms of nutritional content. And at a palatable price too, perhaps £200-300 per 28-day meal plan.

A route forward was emerging but this would mean that Tinkerdash would need to provide the recipes to form the basis of the meal plans, which leads to my next point… I created the recipes.

With the copyright laws regarding the reuse of recipes that are published online somewhat unclear, and not being something I wanted to tackle just yet, I created 20 recipes that could be used for the first meal plan. These were based on meals I’d previously cooked and were vegan, which was the first dietary need that Tinkerdash would aim to help, for reasons discussed in a previous post.

Finally, I transferred the recipe data to a spreadsheet, splitting up the values and units as shown in the image below, as well as converting all measurements to the Metric system. Why did I do this? Simply because I figured it’d be easier for either a human or machine to use in future, by working with this format from the start it might save hours of work down the line. Here’s an example recipe:


There was still work to do to make the process of creating recipes and meal plans repeatable and efficient, however, Tinkerdash did have the basis for the next version of its service.

In order to launch something and get feedback from real users the next question to answer was, how would people prefer to be fed (excuse the pun) this recipes and meal plan information?

Next, the one-month pilot was over and reviews were in. Read what the users had to say here.

Day 150: Why I passed on applying to three start-up accelerator programs

“‘Funded’ doesn’t neccessarily mean ‘validated'” – Wil Schroter

A few weeks ago, Tinkerdash received an unexpected invitation to apply for JLAB, the “the startup accelerator program from the John Lewis Partnership”.

But, flattered as I was, I was torn.

JLAB was now one of a few start-up accelerator programs on my radar. I’d already taken the time to apply (unsuccessfully) for The TrueStart Productivity Challenge in partnership with Morrison’s supermarket in recent months, which was of a similar ilk to JLAB and other programs.

From this, I gathered that the application process for these types of programs typically involves creating a pitch deck, backed up by best-guess financial estimates, followed by rounds of events and meet-and-greets. A time-consuming and somewhat distracting task.

So, in the end, I decided to let the April 2017 application deadlines pass by without applying for three separate accelerator programs.

Sure, I understood the benefits of working with an accelerator program, for example, getting access to investment and mentorship. However, whilst these things would be nice, I felt that the process of applying would detract from Tinkerdash’s top priority to acquire those first 10-20 paying customers, which was the only thing that would truly validate that the current version of Tinkerdash had legs.

And that’s not to say I won’t apply to JLAB or any other accelerator program in future but if I do, I’ll do so when I feel it makes sense for Tinkerdash and its users.

Time will tell if this was a sensible decision. Here’s to hoping!

Next, certain things are best left to the experts. Read about how and why Tinkerdash is trying to work with diet and nutrition professionals here.

P.S. in case you’re wondering which start-up accelerators I was considering applying for, they were:

  1. JLAB, which I’ve already discussed above.
  2. TrueStart, a start-up accelerator program specialising in retail and consumer businesses.
  3. TechStars London, which had been recommended by a friend working for a market leading online investment platform.

Day 140: How I tried to scale my business, starting by standardising it

“Growth is a spiral process, doubling back on itself, reassessing and regrouping.” – Julia Margaret Cameron, British Photographer

In trying to develop a scalable business model, I realised an element of standardisation would be required. At least in the short-term, it would not be feasible to cater for bespoke requirements for every single customer.

This meant Tinkerdash would have to initially target specific demographics and dietary needs but how many people should the standardised meal plans cater for? How many meals per week? What budgets?

To try to answer these questions I began scouring reports from The Office for National Statistics including Families and households in the UK: 2016  and Family spending in the UK: financial year ending March 2016. Here’s what I found out (figures are for 2016 unless stated otherwise):

  • There were 18.9 million families in the UK.
  • There were 12.7 million married or civil partner couple families in the UK in 2016. This was the most common type of family. 7.9 million of these did not have dependent children.
  • 85% of families with dependent children have one or two children. Only 15% have three or more children.
  • In 2013, the average number of dependent children in a family was 1.7.
  • Average weekly household spending remained level at £528.90. £56.80 of this was on food & non-alcoholic drinks and £52.20 of this was on just food.
  • Low-income households continued to spend a higher proportion of their expenditure on food and non-alcoholic drinks (17%, equating to £40.40) when compared with households with a higher income. Households with the highest income spent 7.5% of total expenditure on this category (equating to £72.60).
  • Patterns suggest that lower income households assigned more of their food budget to basic groceries. In contrast, higher-income households spent a higher proportion of their food budget on vegetables.

Based on this, I decided to focus on households of two to four people, which accounted for the vast majority of UK households.

The month-long pilot had shown that people tended to opt for five to seven meals per week and it seemed reasonable to extrapolate this across the population. Based on this, and by anticipating cost-efficiencies by meal planning across four weeks rather than single weeks, it seemed feasible to meet even the tightest of budgets at just £40 per week. It was worth trying at least.

So 3 questions answered. The standardised meal plans would cater for two to four people, eating five or seven meals per week, aiming to be within a budget of £40 per week or £160 for four weeks.

Next, the question of which dietary needs to prioritise.

The pilot included people with a mix of dietary needs including gluten-free, vegetarian, lactose intolerance, plus a load of food likes and dislikes. This time it did not seem reasonable to extrapolate this across the population. Instead, I went back to Facebook adverts to understand representative market sizes of different dietary needs, using the same method I’d written about in a previous post here.

The parameters I used for the search were location “United Kingdom”, age “18-55” and gender “All”. The variable search term was the diet type. These are the terms I searched and the popularity, measured as “people who have expressed an interest in or like Pages related to <search term>”:

  • Veganism – 98.8m people
  • Vegetarian – 96.4m people
  • Healthy diet – 46.2m people
  • Gluten-free diet – 21m people
  • Low-carbohydrate diet – 12.9m people
  • Palaeolithic diet – 12.1m people
  • Diabetic diet – 1.3m people
  • Atkins diet – 1.0m people
  • Lactose intolerance – 997k people
  • Mediterranean diet – 811k people

Based on this, I decided to focus on the top four dietary needs, which were veganism, vegetarianism, healthy and gluten-free.

Google Trends also suggests that “interest” in veganism has been growing at a rate above many other dietary needs since 2014, including the other three listed here.

So, there it was. The next iteration of Tinkerdash would focus on households of two to four people, eating five or seven meals per week, within a budget of £40 per week or £160 for four weeks. And the meal plans would cater for four dietary needs, which were veganism, vegetarianism, healthy and gluten-free.

That is until we learn more and decide to change the direction of Tinkerdash, again!

Day 136: How I prepared to gather feedback from Tinkerdash’s first users

“The art and science of asking questions are the source of all knowledge.” – Thomas Berger

Nearing the end of the four-week review, it was time to start thinking about gathering feedback from the users. To keep things simple, I dropped each of the reviewers a personal note requesting their feedback toward the end of that week.

This would be an opportunity for them to provide open, honest feedback, as well as discussing how we wrap up the review including the publication of any posts on their blogs and social media channels.

In terms of preparation, I constructed a basic questionnaire by reusing some of the questions that had served me well in past, plus some “good” questions as recommended by The Mom Test. The questionnaire concluded with some admin, including whether the blogger would want to continue using Tinkerdash and how much they would pay.

Below show how the first draft looked. Where possible I’d try to ask the question in person or over the phone so to be able to dig deeper into people’s answers.

Before Tinkerdash…
  1. How would you describe your weekly grocery shopping routine? For example, who you shop for, where you typically shop and how.
  2. What are your biggest pains when it comes to grocery shopping? How are you dealing with it now? What else have you tried?
  3. Tinkerdash provides people with their own virtual assistant, who can help you with almost any grocery shopping tasks you wish. If you had access to an assistant, how would you use them? Decided not to repeat this question.
  4. Finally, how would you feel about letting somebody else shop for you?
Whilst using Tinkerdash…
  1. How would you describe your weekly grocery shopping routine?
  2. What are your biggest pains when it comes to grocery shopping? How are you dealing with it now?
Next steps/admin…
  • Do you want to continue using Tinkerdash? Why?
  • If yes, how much would you be willing to pay on a weekly basis?
  • How do you see the review process working?
  • How shall we deliver your financial reward for completing the review?

Finally, is there anything else I should have asked?


With questionnaires going out to reviewers over the course of that next week, it was an exciting time. What would they say? And could their feedback change the course of Tinkerdash forever?

Day 132: The day I realised that my business idea was not scalable, yet

“Profitability is coming from productivity, efficiency, management, austerity, and the way to manage the business.” – Carlos Slim

It was week three of the Tinkerdash review and the feedback from the users was suggesting that the concept had legs.

In fact, one of the users had already asked how much it would cost to continue using Tinkerdash after their review period finishes. Was this to be the first paying customer?

“I know you are not taking customers as of yet but wondered, do you have a price of what the monthly subscription would cost as I want to factor this in? I am really considering keeping you on after I have finished this review as you have been amazing.”

Exciting as this was, Tinkerdash was not yet a scalable business. The time it took to create personalised meal plans alone would be prohibitive, and this was just one of many time-consuming tasks.

So, to try to understand where major change would be required to enable Tinkerdash to scale, I crunched some numbers. Here’s a screenshot of the first draft, based on a number of assumptions explained below:


To explain year 1, it’s assumed that:

  • Tinkerdash would acquire 1000 customers each paying £5 weekly subscription fee.
  • The service would be predominantly powered by human personal assistants, each able to cater to up to 25 customers per week.
  • Tinkerdash would thus require 40 human assistants on its payroll in year 1.
  • The human personal assistants would each cost Tinkerdash around £400 per week.

Based on these figures, even without including the “other” costs such as wages and infrastructure, there would be a loss per customer of £9 each week in year 1. This was concerning.

Fast forwarding to year 3, it’s assumed the service would be more automated and each assistant would be able to cater for up to 100 customers per week. With this, plus rapid growth in user numbers, economies of scale could take effect and Tinkerdash could turn a gross profit. And quite a nice profit at that.

BUT where would the £6,000,000+ required to fund years 1 and 2 come from? Where would the 7500 human personal assistants come from to power Tinkerdash in year 3?

Clearly, there was something broken with this model and it would need to evolve to allow Tinkerdash to grow to the next stage. Challenge set. It was time to start packaging the “best bits” of Tinkerdash into something that was actually scalable!

Next, read about how I got to work creating a scalable business model here.

Day 116: Tinkerdash places its first order, woohoo!

“And… and I’m… I’m real. I’m a real boy!” – Pinocchio

The training wheels are off. Today Tinkerdash placed its first order on behalf of one of its users. Specifically, this involved creating a weekly meal plan and shopping list, reviewing both with the user, and then adding all of the ingredients to their Tesco online shopping basket. The user simply reviewed their basket and checked out, taking all of a few minutes.

What started as just an idea several months ago now feels real. Granted there’s still a mountain to climb but these are the moments we need to enjoy!