How Defining the Customer Problem First Shapes a Startup’s Pitch and Pivots

Andi Boediman, Rektor at Metaguna University, argues that a startup founder’s primary job is to define the customer’s problem with proof, not to build technology or pivot products. The rest of the business model can change up to 90 percent and still succeed, as long as the problem and the founder’s grasp of it remain fixed.

What framework helps a founder explain a startup in seconds?

Boediman meets 200 to 300 founders each year, spending Wednesdays from 9 a.m. to 6 p.m. in pitch meetings. He then selects only 5 to 7 companies annually for investment from a portfolio of 30 he has built since running a venture fund that managed 12.5 million and grew nearly four times in value.

He says most founders cannot state their idea briefly.

"Usually even if you have a good business, when asked what your product is, you need 15 minutes to explain. That's too long; you must be able to convey it in a very short time."

His remedy is a five-part breakdown that forces clarity before any slide deck.

Brand, product, customer, problem, defensibility

The first component is brand. The name may come from a spouse, a partner, or even a divine hint.

"You can choose your wife’s name, girlfriend’s name, whatever, inspiration from heaven is all allowed."

A name like Apple is not descriptive yet becomes functional through use. The second is product: it must be simple and clear, whether a physical item or a service. If the product description is complex, the listener cannot judge it quickly.

Third is customer. The customer base must be large and growing, and the founder must have a way to measure that growth. Boediman’s own fund returned multiples because portfolio companies reached scale, so he screens for that trajectory. Fourth is problem. The problem must be large and must demand solution now.

He illustrates with a restaurant near a university.

"If you open a restaurant, that does not solve an immediate problem, because tomorrow others will open next door. The problem is not specific, not unique."

Fifth is defensibility. The startup needs something others lack, making it unique and hard to copy. Without this, a large market only invites copycats who erase margin.

Which element matters most

Among the five, he singles out one as decisive.

"From all of these, which one is the most important? I suggest: the problem."

This hierarchy lets a founder ignore polish elsewhere if the problem is sharp. The other four elements support the pitch, but the problem is the anchor that decides investment.

Why is defining the problem the founder’s key skill?

Boediman challenges the myth that technical brilliance builds startups. He lists three founder types that are not enough.

"A good founder is not one who is good at technology, not one who is good at making products, not one who is good at selling, but one who can define the problem."

The definition must survive contact with reality. He looks for problems he had not imagined, shown by numbers.

"I never thought there was that problem, and it is proven with data."

Therefore his capital rests on two pillars only.

"When I invest in a startup, I only bet on two things: the problem and the founder."

The product, customer segment, or sales trick can shift; those bets do not move. This is why he can accept massive pivots later without re-evaluating the core thesis.

How did the fish-feed device reveal the real problem?

He backs this with a fish-feed startup he funded when others refused. The founder showed a massive Indonesian aquaculture market and a clear cost split. Feed expense is 80% to 90% of total farming cost. Trimming 10% of feed cost lands directly as profit.

"If I can save just 10% on feed cost, that immediately becomes profit. Does that make sense? Yes, it makes sense."

The founder learned this by managing ponds in college. The deeper issue was overfeeding: excess food fouls water, shifts acidity, triggers disease. That root cause is what the startup needed to attack.

The automatic feeder and its weak moat

The startup built a sensor-driven auto feeder. Boediman tested its durability by imagining a copycat in China.

"If you bring this to China, how long before it's copied? Someone buys the bottle next door, has the chip, and assembles a prototype in a week. We are done."

Hardware alone was not defensible. Yet it solved a human gap: pond owners and daily feeders are different people. Feeders skip shifts, overfeed, and waste stock. The device forced correct dosing. Priced at tens of millions, though, no farmer bought it.

Why farmers used it without buying

The founder camped at farms until sympathy won.

"The answer was not skill, but pity: 'I feel sorry for this guy hanging around every day.'"

Within a month, feed use dropped and savings appeared. That proof turned the device into a trust marker. Feed suppliers then asked to ride the channel.

"After seeing this, the feed companies said: it's nice to sell through this startup, you sell my goods through them."

How did the startup pivot to feed distribution and financing?

The proof of savings changed the revenue model. The firm stopped relying on device sales and earned from feed itself. The step-by-step path shows how a problem-first mindset expands:

1. Founder identifies large aquaculture market and feed cost share of 80% to 90%.
2. Root cause found: overfeeding from split owner/feeder roles.
3. Automatic feeder built, but hardware not sold due to price and copy risk.
4. Free use via founder’s presence yields usage data showing 10% saving.
5. Feed companies partner to distribute sacks through the device.
6. Device lent free; profit taken from recurring feed sales with long lifetime value.
7. Feed makers mandate device use to lock farmer brand loyalty.
8. Farmers want expansion capital; banks refuse without data.
9. Startup’s pond data becomes validator for cooperatives to lend.
10. Loans issued in feed not cash, ensuring on-farm use.
11. Free website aggregates buyer demand, standardizes fish size.
12. Startup becomes off-taker (middleman) at 100,000 points nationwide.

"There is nothing special about it, but as long as they farm, they always buy from him; lifetime value is very long."

Becoming a channel for credit without planning fintech

Farmers wanted new ponds but banks dismissed them. The startup held the only pond yield data.

"All cooperatives said: I trust this startup to channel debt, because the business is correct."

Loans were paid in feed, not cash, so funds stayed in farming. The firm became a fintech by effect, not intent.

"Fintech is not a goal; fintech is a business model."

Scaling to off-taker and standardizer

A free website let buyers pool orders. Fish sizes were standardized, and the startup took the role of middleman.

"In the village language, it's a broker; in crude terms, a tengkulak."

Now present at 100,000 points across Indonesia, it supplies device, credit, and buyback. The original feeding problem stayed constant while the business form rotated.

Why can 90% of the model pivot freely?

Boediman closes with the permission to mutate. Once problem and customer are locked, form is flexible.

"90% of your business model can pivot, it doesn't matter. The rest, all those bullshit, doesn't matter anymore."

The founder’s worth is the accuracy of problem definition, not the first prototype.

"Startup is a mindset: you define the problem. That is the only two things that matter when I invest."

The fish-feed company morphed from hardware maker to feed shop to lender to off-taker, yet never redefined its core problem of feed waste and lazy feeding.

Key Takeaways

• A founder must define the customer’s problem with data; that is the only investment criterion alongside the founder.
• Explain a startup in seconds using five parts: brand, product, customer, problem, defensibility—with problem as the core.
• The fish-feed case shows hardware can be copied, but the feeding problem and data defensibility create lasting business.
• Pivot product, sales, or model freely; the problem and founder fit are the only bet an investor needs.
• Scale by following the problem into adjacent services (feed, credit, off-take) without changing the core problem.