How to Choose a Business Idea With £600: Julie Deane's Scoring Decision
Short answer: Choose an idea that can reach a paying customer within six weeks for no more than £300, leaving half your £600 uncommitted for fulfilment, refunds and a second test. Score each idea on customer access, time to payment, cash exposure, contribution per sale, personal advantage and reversibility. Reject any idea that needs premises, a large production run or paid advertising before you can collect evidence.
The question is not which business sounds most exciting. With £600, your first job is to buy evidence without losing the ability to act on it.
My view is that a service or made-to-order offer should beat an inventory idea unless the product can be pre-sold or sourced in tiny quantities. Limited capital is not a disadvantage you overcome with confidence. It is a design constraint you use to eliminate slow, cash-hungry options.
Use the Six-Week Idea Scorecard
The Six-Week Idea Scorecard gives each idea zero, one or two points across six tests. Score from evidence you can obtain now. Do not award points because you believe you will become good at something later.
| Test | 0 points | 1 point | 2 points | |---|---|---|---| | Customer access | You cannot name where buyers are | You can reach a broad audience | You can contact 20 likely buyers directly | | Time to payment | More than six weeks | Three to six weeks | Within two weeks | | Cash before proof | More than £300 | £101 to £300 | £100 or less | | Contribution | Unknown or negative | Positive but below £50 per sale | At least £50 per sale or 50% of price | | Personal advantage | No relevant access or ability | Transferable skill | Direct skill, credibility or supplier access | | Reversibility | Lease, debt or unusable stock | Some recoverable equipment | Can stop with little further cost |
Reject any idea scoring zero for time to payment, cash before proof or reversibility. Then test the highest score. A score chooses the next experiment, not your identity for the next decade.
The contribution threshold is a working guide for a £600 start, not a universal rule. A lower-value repeat service can still work if acquisition and delivery costs are low and retention is observed.
What Julie Deane's spreadsheet actually established
Julie Deane needed a way to fund different schooling for her children. In her own account of starting the Cambridge Satchel Company, she says she had £600 in savings, listed 10 possible businesses and scored them on factors including required funding and time until a product could be sold. Satchels ranked first.
In a Sage interview, Deane describes having the school summer holidays to start. The six-week test here is your working timeframe, not her documented schedule. Her £600 funded a beginning, not the mature company. She found a manufacturer, built a website and won an initial bulk order of 10 bags.
The later scale does not prove that satchels would win your scorecard. Deane had observed a specific problem, could describe the product and found a route to small-batch supply. Your best idea may be bookkeeping, garden maintenance or product photography.
The transferable decision is to compare ideas against the constraint that matters, then commit to one test. Endless ideation would have consumed the deadline without improving the evidence.
Convert your personal target into a business constraint
Write the amount you need, the date you need it and whether that amount means revenue, contribution or personal income. Those are not interchangeable.
If you need £1,200 a month for living costs, a business generating £1,200 in sales is insufficient. It must also cover materials, payment fees, travel, insurance, tax and cash retained for the next cycle. Calculate the required contribution first.
Then list your practical constraints: hours available, transport, workspace, caring responsibilities, contractual restrictions and skills you can sell now. Do not assume you can learn a regulated trade, obtain permissions and build customer trust inside six weeks.
Anything touching food, children, health, finance, construction, transport or personal data may require specific licences, insurance or qualifications. Requirements vary by country and sector. Check the relevant authority and obtain qualified local advice before accepting money.
Generate ideas from assets and expensive problems
Make three short inventories. First, list tasks you can already perform to a paid standard. Second, list people and organisations you can reach without advertising. Third, list problems those people already spend money, time or risk managing.
Combine one item from each inventory. If you can photograph products, know independent retailers and observe that their online listings are inconsistent, a fixed product-photography day is testable. “Start a creative agency” is too broad to score.
Keep the ideas at offer level. “Pet business” cannot be priced or tested. “Saturday dog walks for hospital staff within two miles” can.
Aim for 10 ideas, then remove any that rely on an unverified assumption about licensing, supply or customer access. Research should make the list smaller.
Worked example: Elm & Buckle Bags
In this fictional example, Priya has £600 and six weeks. She can sew and knows independent bookshops. Compare four offers:
| Idea | Access | Speed | Cash | Contribution | Advantage | Reversibility | Total | |---|---:|---:|---:|---:|---:|---:|---:| | Made-to-order canvas book bags | 2 | 2 | 2 | 2 | 2 | 2 | 12 | | Imported leather handbag range | 1 | 0 | 0 | 1 | 1 | 0 | 3 | | Bag-repair collection service | 2 | 2 | 2 | 2 | 2 | 2 | 12 | | Weekend sewing workshops | 1 | 1 | 1 | 2 | 2 | 1 | 8 |
Repairs and made-to-order bags tie. Priya commits £286 to two tests:
| Test cost | Calculation | Amount | |---|---:|---:| | Repair materials for 10 jobs | 10 × £7 | £70 | | Public-liability and equipment adjustment | Short-term attributable cost | £86 | | Three book-bag samples | 3 × £28 | £84 | | Product photographs and local delivery | Direct cash cost | £46 | | Initial cash used | £70 + £86 + £84 + £46 | £286 |
Priya charges £48 per repair and £62 per bag, taking £31 bag deposits. In three weeks, she completes eight repairs and four bags, collects every balance and pays £112 to produce the ordered bags. Samples remain separate testing costs.
Repair contribution is 8 × (£48 - £7) = £328. Bags contribute 4 × (£62 - £28) = £136. Together they produce £464 before labour and testing costs.
Repairs take 10 hours and bags eight. At £18 an hour, labour is £324. Contribution after labour is £140, but deduct the £216 testing costs (£86 + £84 + £46): the trial loses £76. Cash remaining before drawings is £600 - £286 - £112 + £384 + £248 = £834, with £14 of unused repair materials.
Repairs earn £32.80 per labour hour before paying Priya, versus £17 for bags. She starts with repairs and adjusts bag pricing before retesting. Paid demand has not made both offers viable.
These figures are illustrative. Use your own labour value, insurance, tax and direct costs.
Spend in evidence-sized steps
Divide the six weeks into three releases. Spend no more than £100 to secure conversations, supplier facts or a crude sample. Release the next £200 only when at least five suitable buyers confirm the problem and two agree to pay or place refundable deposits. Preserve the final £300 until you have to fulfil real orders.
This is not timidity. It prevents branding, bulk materials or a website from consuming the money before a customer makes a decision.
Deposits and pre-orders must be described accurately, with delivery dates and refund terms you can honour. Consumer rights, tax and distance-selling requirements vary by jurisdiction and product. Use qualified local advice where the obligation is unclear.
Choose one idea after the test
At the end of week six, compare collected revenue, contribution after founder labour, acquisition effort, repeat intent and remaining cash. Choose the offer with the strongest paid evidence and a delivery model you can repeat.
Do not keep three offers alive because each received one compliment. Focus is valuable when capital is limited. You can revisit the second-ranked idea after the first produces reliable cash or clearly fails.
Also allow a stop decision. If nobody pays after you have reached 20 suitable buyers with a specific offer, do not spend the remaining cash polishing it. Review the segment, problem and price, then run a different test.
What to do over the next six weeks
In week one, define your income target and list 10 narrow offers. Research permissions and score each idea. During week two, speak to five likely buyers for the top two ideas and obtain supplier or delivery costs.
By the end of week three, present a paid test to at least 20 suitable buyers. Use weeks four and five to deliver the smallest honest version and record every hour and direct cost. At the end of week six, rank the tested ideas on contribution, repeat demand and cash remaining. Commit to one for the next 90 days or stop without spending the reserve.
Related guides
Frequently asked questions
What business can I realistically start with £600?
Usually, choose a service, repair, brokerage or made-to-order product that uses skills and equipment you already have. Examples might include bookkeeping support, product photography, garden maintenance, alterations or a narrowly defined local service, subject to your ability and local rules. £600 is rarely enough for premises, broad inventory or a regulated operation. The amount should fund customer evidence and first delivery, not a polished launch. Your location, skill, insurance and customer access determine the answer, so score actual offers rather than copying a list of cheap businesses.
Should I choose the idea with the highest profit margin?
No. Margin without reachable customers or timely payment does not solve a six-week constraint. Compare contribution per sale, hours required, acquisition effort and cash timing. A 75% margin product that takes three months to source may be inferior to a 45% margin service paid this week. Also check whether the margin survives returns, delivery and your labour. Use the scorecard to balance economics with access and reversibility. Once two ideas pass the constraints, paid demand should decide between them rather than theoretical margin alone.
How many people should I interview before choosing?
Speak to at least five suitable buyers for each of your two leading ideas, then put a paid offer in front of 20 people in the stronger segment. Interviews reveal language, existing alternatives and objections, but they do not prove purchase. Ask about recent behaviour rather than whether the idea sounds good. If the buyers describe the problem but nobody accepts a specific paid next step, treat demand as unproven. A niche B2B offer may require fewer, deeper conversations; a low-value consumer product may require more behavioural evidence.
Do I need a website before taking the first order?
Usually, no. You need a clear written offer, a lawful way to take payment, accurate contact details and any required consumer information. A simple page may help when buyers need to verify you, but building a full site before testing payment can consume the budget. Use the channel your intended customer already trusts, then add a website when it removes a demonstrated sales or service obstacle. Certain ecommerce, privacy and consumer-information duties apply regardless of site size, so check current local requirements rather than treating informality as exemption.
Should I borrow more if the best idea needs over £600?
Not before separating proof cost from scale cost. Find the smallest version that tests whether customers pay, such as a paid pilot, sample, deposit or manual service. Borrowing may be reasonable after demand and contribution are evidenced and repayment fits a downside cash forecast. It is dangerous when the loan funds an untested assumption or a minimum order you cannot resell. Credit terms, security and personal liability vary, so obtain qualified financial and legal advice. Limited money should make the first test smaller, not automatically make the debt larger.
What if two ideas receive the same score?
Choose the one that can produce paid evidence sooner and leave more of the £600 recoverable. If those are equal, run parallel micro-tests capped at £100 each, using the same period and comparable customer quality. Decide from collected payment, contribution after labour and repeat intent. Do not modify the score after seeing which idea you personally prefer. A tie means the desk research has reached its limit. Customer behaviour should now break it. Stop one test once the result is sufficiently clear and preserve the remaining cash.
How do I value my own time when I am not paying myself?
Use a realistic replacement hourly cost or the amount you could earn from available paid work. Record it separately from cash so you can see both immediate survival and sustainable economics. If a £60 sale requires four hours and £12 of materials, it creates £48 of cash contribution but only £8 after valuing your time at £10 an hour. That distinction prevents you choosing an idea that works only through unpaid labour. Your initial rate can be indicative, but update it as you learn how long delivery actually takes.
Comments (0)