Target customer profile: from “we can do anything” to three verifiable traits

Bottom line first:“We can do anything” means you have no target market, and your content, channels, and pricing will all lose focus. A usable ICP doesn’t have to be complex. It needs only three things: verifiable evidence of size, verifiable buying-behavior signals, and a decision chain you can actually reach. Start by reverse-engineering your last 20 closed customers, write your exclusion list before your target list, and judge whether the profile holds up by your inquiry match rate and quote-to-close rate.

1. “We can do anything” isn’t a strength. It means you have no target market.

Many companies going global introduce themselves like this: “We serve manufacturing clients worldwide, and our products fit every scenario.” It sounds ambitious. In practice, it means you do nothing well.

The problem isn’t mindset. It’s mechanics. Your target customer profile determines three things: what content you write, which channels you invest in, and what your quote looks like. When those three don’t point the same direction, your effort scatters.

Project	Clear ICP	No clear ICP	Source of the gap
Content topic hit rate	60%–80% relevant to the core business	30%–50%	Do topics come from customer problems or from your imagination?
Inquiry match rate	40%–60%	15%–25%	Are the people coming in the people you want?
Quote-to-close rate	20%–35%	8%–15%	Does your quote structure address what the customer cares about?
Sales cycle	30–60 days	60–120 days	Do you reach the decision chain in the first round?
Follow-up cost per customer	¥800–1,500	¥2,000–4,000	Share of wasted conversations

The biggest gap isn’t inquiry volume. It’s that follow-up cost per inquiry differs by 2 to 3 times. Because you keep explaining the same thing over and over to people who were never going to buy.

One boundary needs stating here:a customer profile is not better the narrower it gets. Serving only a profile like “Germany, annual procurement over ¥5 million, automotive Tier 1” will starve your business if you’re under ¥10 million in annual revenue. The right width is this: narrow enough that you can write targeted content, wide enough to keep a customer pool 3 to 5 times your current capacity.

2. Four groups of ICP fields, each with only 3 to 5 entries

The easiest way to go wrong with a profile is to write dozens of fields that nobody uses. In practice, four groups are enough.

Group	Field	Example entry	Purpose
Company attributes	Industry segment	Industrial automation equipment manufacturing	Determines your content topics
Company attributes	Size tier	Annual revenue ¥200 million to ¥1 billion	Determines expected deal size
Company attributes	Region	Germany, Italy, North America	Determines language and time zone
Company attributes	Channel structure	Mostly direct sales, with local service providers	Determines how you reach them
Company attributes	Product complexity	Many models, high degree of customization	Determines how you quote

Group	Field	Example entry	Purpose
Buying behavior	Purchase frequency	2–4 batches a year	Drives inventory planning and payment terms
Buying behavior	Decision cycle	2–5 months	Drives follow-up cadence
Buying behavior	Decision-making role	Signed off by both engineering and procurement	Drives who sees your content
Buying behavior	Willingness to switch suppliers	Open to a second supplier	Drives your opening pitch
Buying behavior	Certification requirements	Needs CE or industry certification	Drives what materials you prepare

Group	Field	Example entry	Purpose
Pain point intensity	Main pain points	Unstable lead times, high defect rates	Drives your content topics
Pain point intensity	Pain point in dollar terms	An hour of downtime costs tens of thousands	Drives your value proposition
Pain point intensity	How they solve it today	Hedge with inventory	Drives your entry point
Reachability	Publicly listed contact roles	Procurement manager, technical director	Drives who you target
Reachability	Reachable channels	Trade shows, industry directories, LinkedIn	Drives your channel mix
Reachability	Decision-making language	English, German	Drives which language versions you need

Of the four groups of fields,the fourth, “reachability,” is what really decides success or failure. The first two tell you what your customer looks like, the third tells you what to say, and the fourth decides whether you can actually reach them. If your profile turns up no publicly reachable role or channel, it stays on paper.

3. Three verifiable traits: whether a customer is your customer comes down to just these three

Once the profile fields are filled in, the next step is to compress them into three verifiable judgments. Verifiable means:you can confirm or rule it out in 15 minutes using public information, not by gut feel.

Characteristics	Specific meaning	How to verify	Disqualifying signal
Verifiable scale	Publicly verifiable business scale	Website, annual report, hiring volume, plant size	Website is a single page with no scale information at all
Verifiable buying activity	Signs of relevant buying activity happening now	Job postings, certification announcements, trade show records, number of SKUs	No news for three years, no job postings
Reachable decision chain	You can find a specific role or channel entry point	Contact page, LinkedIn roles, trade show directory	Only an anonymous form, no role information

Verifiable signal	Where to look	What it tells you
Job postings	Job boards, company careers page	Hiring mechanical engineers means they’re expanding or developing new products
Certification announcements	Certification bodies and industry directories	Industry certification means they’re serious about this business
Trade show participation records	Trade show exhibitor directories	They invest in marketing and are an active buyer
Channel structure	The “dealers” or “sales network” page on their website	Tells you whether they sell direct or rely on channels
Number of SKUs	Number of product pages and how often they’re updated	Reflects product complexity and iteration pace
Number of website languages	Language switcher on their website	Reflects whether they buy and sell across borders
Supplier page	The “suppliers” or “procurement” entry on their website	Directly exposes the procurement entry point and requirements

The value of these three traits is that they are falsifiable. Write “the customer values quality” and you can never disprove it; write “the website has 3+ languages and relevant job postings in the past year” and you can tell true from false in 15 minutes.Only a profile that can be falsified has screening power.

4. Five moves to reverse-engineer your ICP from closed customers

Building a persona out of thin air is the easiest way to get it wrong. The right order is to work backward from customers who have already bought.

Steps	Action	Time	Deliverables
1	Break down your last 20 closed deals	2 to 3 days	Fill in one four-part form per customer
2	Extract common traits	1 day	The 5 to 8 traits that appear most often
3	List exclusions	Half a day	The three types of customers you don’t serve
4	Rank by priority	Half a day	1 primary persona, 1 to 2 secondary personas
5	Backfill the fields into your CRM	1 to 2 days	A filterable customer database

Step one is where people cut the most corners. Many teams sum it up from memory: “our customers are mainly mid-to-large manufacturers in Europe and North America.” But pull up your last 20 deals and often half are from Southeast Asia, with order values a third of the European ones.

When you extract common traits in step two, your criteria need to be specific:A trait has to appear in at least 12 of the last 20 deals to count as common. Traits appearing 5 to 11 times go into a secondary persona; anything under 4 stays out of the persona entirely. The threshold isn’t precise, but it’s useful, and it stops you from mistaking a coincidence for a pattern.

Steps three through five are the ones people skip most:Exclusions and priorities have to live in CRM fields, otherwise your persona is just a slide deck. At minimum, your CRM needs to filter on four dimensions: industry segment, company size, region, and decision-making role.

5. How three typical personas compare, and how each prefers to be reached

These are all manufacturing customers, but what they care about differs a lot. Three types come up most often:

Dimension	Technology-driven	Cost-driven	Project integration
Typical industries	Precision equipment, automation	General hardware, consumables	Engineering firms, system integrators
Average order value	¥100,000 to ¥1,000,000	¥10,000 to ¥100,000	¥300,000 to ¥3,000,000
Decision-making role	Engineering leads, procurement supports	Procurement leads	Project lead leads
Decision cycle	3 to 8 months	1 to 2 months	4 to 12 months
Core concerns	Specs, stability, traceability	Unit price, payment terms, lead time	Delivery schedule, interface coordination
Content preferences	Technical white papers, spec comparisons, test data	Spec sheets, price ranges, inventory and lead times	Case studies, project milestones, delivery plans
Effective channels	LinkedIn, industry trade shows, technical communities	Marketplaces, industry directories, email	Trade shows, channel partners, case study pages on your site
Common mistakes	Pitching a technology buyer with price talk	Pitching a procurement buyer with long technical write-ups	Pitching a project buyer with a standard quote

The mistake people make most often across all three personas is using the same content to pitch all three. Technology-driven customers see a page full of “price advantage” and skip right past it; cost-driven customers read three thousand words of technical analysis and think you’re dodging the question. A simple test is to look at the first question a customer asks in their inquiry: specs and test reports means technology-driven, price and lead time means cost-driven, project schedule and how you’ll coordinate means project-driven.

Take technology-driven customers. Their path is usually technical screening first, then commercial comparison, so your content needs to come earlier. For more on how these buyers research, see How overseas buyers research breakdown.

6. An exclusion list matters more than a target list: three types of customers you don’t serve

A target list tells the team where to go. An exclusion list tells the team where not to go. The exclusion list is more actionable because it saves time directly.

Customer type you don’t serve	Warning sign	Why you turn it down	How to respond
Only wants the lowest price	Asks for your floor price in the first email, never discusses specs	Paper-thin margin after the deal closes, high after-sales cost	Give standard MOQ and a price range
Needs beyond what you can deliver	Asks for a category unrelated to your core business	You can’t guarantee delivery quality, and your reputation takes the hit	State your limits clearly and refer them to a peer
Can’t reach the decision-maker	No way to reach anyone in a decision-making role	High communication cost, very low conversion	Send materials only, no custom work

Exclusion	Test	What to do
Order value below the economic line for your capacity	Below minimum order quantity with no sign of growth	Send the catalog only, no quote
One-off price shopping	Refuses to share project details twice in a row	Move to the low-frequency outreach pool
Demands payment terms far beyond the norm	Asks for long payment terms on the first order with no guarantee	Decide after a risk assessment

When you write an exclusion list, guard against one side effect:sales teams can easily use it as “an excuse to turn customers away.” So write it as rules you can judge against, not as an attitude. Don’t write “no small customers.” Write “if a single order falls below the minimum order quantity and there’s been no repeat purchase in 12 months, send the catalog only and no quote.”

7. Turning the profile into content and ad spend, and how often to review

The profile has to end up as an executable table, or it’s just a document.

Profile	Content topic	Main channel	Conversion path	Primary metric
Technology-driven	Spec comparisons, test methods, certifications and compliance	LinkedIn, trade media, industry trade shows	Technical document download to technical discussion	Technical document downloads
Cost-driven	Spec sheets, price ranges, lead times and inventory	Marketplaces, industry directories, email	Inquiry to fast quote	Inquiry-to-quote rate
Project integration	Delivery case studies, project milestones, interface solutions	Trade shows, channel partners, case study pages on your site	Case study read to solution discussion	Time on case study pages and conversion

To judge whether a profile holds up, don’t look at how much you wrote. Look at whether three metrics improved:

Metrics	Observation period	Reference threshold	What to do if you miss it
Inquiry match rate	Quarterly	Above 40%	Adjust content topics and channels
Quote conversion rate	Quarterly	20% or more	Adjust quote structure and entry point
Sales cycle	Every six months	Within 60 days	Check whether you’re reaching the decision-making chain
Follow-up cost per customer	Quarterly	Below ¥1,500	Tighten your screening criteria

The review cadence is check metrics every quarter, revise the profile every six months, based on shifts in the mix of customers who close: if three in ten new customers don’t fit the current profile, it’s time to change the profile.

When you should deliberately loosen the profile:

Scenario	How to loosen it	Duration	Metrics to watch
New market trial period	Drop the size threshold for now and take everyone	3 to 6 months	Inquiry volume and source mix
When you have spare capacity	Take low-ticket orders to fill the gaps	By capacity cycle	Marginal gross margin and equipment utilization
New product validation period	Take cross-industry customers for trials	3 months	Repeat purchase rate and product feedback

Loosening doesn’t mean abandoning the profile,Its purpose is to collect samples; once you have enough, tighten it again. The most common mistake is loosening it and forgetting to tighten it back, so six months later the team is back to “we do everything.”

Once the profile is set, your sample and prototyping policy is the most direct next step. Check it against the four quadrants in Sampling and Proofing Policy: Should You Charge for Samples, and How to Charge Without Killing Conversions; and for how to write customer cases that your target profile actually understands,How to write customer cases that piece gives a six-part structure.

FAQ

Our customers are all over the place. Should we narrow down?

Yes, but not by cutting everyone off right now. Start with two moves: pull apart your last 20 closed customers and look for what they have in common; then work out the follow-up cost and gross margin per customer for each type, and let the data show you which ones are worth keeping. After that, put your effort into the “high margin, high fit” ones first and keep the rest on low-cost outreach. Outright rejecting your existing customer base is very risky, especially when you are not sure which type can carry next year’s revenue.

How do we verify an overseas customer’s real size?

Cross-check using public information, and only trust it when at least two sources agree. There are four common sources: the company’s About and factory pages (look at plant area, headcount, year founded), the number and type of open roles on job platforms, industry directories and association member lists, and local business registration lookups. A single source is easy to inflate, for example a website that only says “global leader” without any size figures cannot be used as a basis for judgment.

Should profiles be split by country?

Yes, but do not split them too finely by country. Country differences affect three things: language version, certification requirements, and time zone and communication habits. If these three things do not differ much in your category, split by region (Western Europe, North America, Southeast Asia) instead of building a dozen country profiles from the start. The proven approach is to define your main profile by “industry plus size,” treat region as a sub-dimension, and adjust only language and certification for each region.

Existing customers are the bulk of our business. Should we still look for new profiles?

Yes, but be clear about why. If existing customers are more than 70% of your business and orders are stable, the goal of finding new profiles is not to replace them but to reduce concentration risk, because when a single customer is too large a share, losing them once can leave 30% of your capacity idle. Split your resources into two tiers: 70% to maintaining your existing customer structure and repeat orders, 30% to testing one new profile, with a 6-month test period, and at the end decide whether to continue based on inquiry match rate.

If the profile and the product positioning conflict, which one wins?

Follow the customer’s buying behavior. Product positioning is about what you want to be; the customer profile is about why customers buy from you. When the two conflict, first look at which value point customers are actually buying in your closed orders. The way to do it is to list the reasons behind your last 20 deals. If more than 70% bought because of lead time and reliability, while your positioning says “great value,” then what needs to change is the positioning copy, not the customer.

How often should a customer profile be updated?

The usual rhythm is a light revision every six months and a full rebuild once a year. The light revision is driven by quarterly metrics (inquiry match rate, quote conversion rate, deal cycle) and only touches fields and thresholds. A rebuild means going through the breakdown of the last 20 deals again, because both your customer mix and your product line can shift within a year. One more trigger: if more than 30% of newly closed customers don’t fit the current profile, don’t wait for the six-month mark — review it right away.

Final thoughts

The value of a customer profile isn’t how good it looks on paper. It’s whether it lets your team reach the same call: whether to pursue this inquiry, whether to send a free sample, whether to write this piece of content.

Where to start: spend two or three days breaking down your last 20 closed customers and filling in the fields, pull out the common traits, then write an exclusion list. Next, narrow it to three verifiable traits and map them onto four filter dimensions in your CRM. After that, check your inquiry match rate once a quarter. You don’t need any software for this — a shared spreadsheet is enough to begin.

The boundaries of a profile are clear:it stops working as your market, products and capacity change, so it’s never a one-and-done task. If you want to build this filter into your website’s forms and content structure so mismatched inquiries are screened out before they reach you, go to /contact/ and tell us about your category and where you stand.

One line worth remembering: a good customer profile doesn’t tell you who to serve — it lets you say in under a minute who you won’t serve this time.