1. One brief, three quotations
A Shenzhen medical device distributor wanted a new logo. The founder asked marketing to get three quotes.
The first came back at RMB 3,800 (roughly USD 530): three concepts, five working days. The second came back at RMB 46,000 (roughly USD 6,400): research and presentation included, six weeks. The third did not quote at all. It asked a series of questions instead — where will this be used, are you planning overseas markets in the next three years, how many product lines do you have now, will that number grow, have you selected your trademark classes, who signs off? Then it said: what you need may not be a logo.
The marketing lead told me later that at the time it felt like posturing. Six months on he had changed his view. The company launched a Southeast Asia business and needed an English brand name and a Latin-script version of the mark — at which point they discovered the new Chinese-character logo barely extended into a Latin environment. Around the same time, the trademark application was rejected once, on grounds of similarity to a prior mark.
"The expensive thing wasn't the RMB 46,000. It was those six months."
The lesson is not that expensive is better. It is this: logo quotations span from a few hundred to several hundred thousand RMB, and the difference is not in how well the mark is drawn. It is in what the mark is being asked to carry.
A mark that hangs on an office wall and sits on a business card, and a mark that must enter dozens of touchpoints over ten years, support several product lines, survive trademark examination, and hold up in both Chinese and English — these are two different engineering problems. Purchasing them with the same word guarantees the thirty-thousand-fold price confusion.
This article sets out where the line falls.
2. Four dividing lines: which side is your brief on?
Use these four to determine whether a logo brief is a single-deliverable purchase or a systems project. Two or more falling on the right means it should not be procured as a single deliverable.
Dividing lineSingle deliverableBrand systemTouchpoint countUnder ten: card, signage, social account, contractsDozens to hundreds: multiple product lines, channels, media, languagesDecision chainOne or two people sign offMulti-department review; board or upward reporting requiredLifespanOne to three years; replaceableTen years or more; replacement is costlyFunctions carriedRecognition onlyTrademark registrability, cross-language adaptation, cross-media stability, systemic extensibilityLine by line
Touchpoint count. The easiest to quantify. List every place the mark will definitely appear over the next two years. If the list runs past twenty rows, what you need is not a graphic but an identity system that holds under varying conditions.
Under ten touchpoints, the mark's primary form suffices. Past twenty, you need its variation rules — horizontal and vertical lockups, single-colour and reversed, minimum size, light and dark grounds, motion contexts, the circular crop that social platforms impose. These are not bonus attachments; they are the conditions under which a mark survives real environments.
Decision chain. A project signed off by one person can proceed on aesthetic preference. A project reviewed by many must proceed on explicable reasoning.
The practical implication: the longer the decision chain, the higher the value of upfront research. A presentation must satisfy not one person's taste but a group's separate concerns. Without a basis for judgment, proposals get revised indefinitely and end up as compromises nobody objects to and nobody endorses.
Lifespan. This line determines whether you can afford to experiment.
A mark you expect to redo in three years can take stylistic risks. A mark meant to last ten must avoid every short-cycle aesthetic trend. The generation of technology logos built a decade ago on gradients, highlights, and skeuomorphic texture has largely been redone since. The longer the lifespan, the greater the need for restraint.
Functions carried. The most frequently overlooked line, and the one that causes the most rework. Sections 4 and 5 develop two of its components — trademark registrability and replacement cost — the two most expensive hidden invoices in this decision.
3. Five market tiers: what each actually buys
Logo pricing distributes across roughly five tiers. What follows is not a price list. It is an account of how the working method at each tier determines what you can receive. Understanding the method is more useful than memorizing the price.
Tier 1 — Templates and AI generation
Method: select a template or enter a name in an online tool, or generate with AI and adjust. Minutes to hours.
What you get: a usable graphic file.
What you do not get: uniqueness (a template may already be in use by hundreds of companies), trademark registrability (near-identical templates carry high rejection probability), variation rules, or judgment of any kind.
Where it is appropriate: internal projects, temporary events, early-stage products still in validation, community and personal brands. In those contexts it is a sound choice and does not deserve disparagement.
Where it is not: a company mark intended for long-term operation, trademark registration, or formal commercial use.
Tier 2 — Independent designers and small studios
Method: one or two people, direct communication, short cycles, typically two or three directions with one or two revision rounds.
What you get: an original mark, basic specification files, efficient direct communication.
What you typically do not get: systematic sector research, trademark risk pre-screening, multi-language considerations, long-cycle extension rules.
Where it is appropriate: SMEs with limited touchpoints, short decision chains, and moderate lifespans. Supply at this tier in Shenzhen is deep, and includes independent designers of genuinely high calibre.
How to screen: look at consistency across the portfolio rather than the brilliance of any single piece — stability is where independent designers differ most.
Tier 3 — Mid-sized design firms
Method: an established project process, presentation and specification production, typically some upfront research, four to eight weeks.
What you get: an original mark, a coherent base specification (colour, type, lockups, misuse examples), and a set of applications.
What you may not get, depending on the firm: deep strategic research, trademark legal support, cross-language and export adaptation, systems-level extension rules.
Where it is appropriate: growth-stage companies with multiple touchpoints needing specification support, but without complex business structures.
Tier 4 — Professional brand agencies
Method: strategy research up front, multiple presentation and review rounds, a complete specification system and application suite, eight to sixteen weeks.
What you get: a mark grounded in judgment, a full visual identity system, executable application standards, and usually trademark registrability considered up front.
Where it is appropriate: companies with many touchpoints, long decision chains, long lifespans, and a need to explain the outcome to a board or investors.
Tier 5 — Systems firms
Method: the mark is designed as one interface within a system; the weight of delivery sits in rules, parameters, and governance mechanisms.
What you get: beyond mark and specification — a parameterized visual system, template architecture, boundaries for AI generation, cross-language semantics, and a maintenance mechanism for after the project ends.
Where it is appropriate: multi-product-line groups, companies going global, companies preparing for IPO, or organizations already producing content at scale with AI.
An honest note on price
Comparing across tiers is meaningless. Tier 1 and Tier 5 solve different problems, in the way a bicycle and a freight truck cannot be compared on cost per kilometre.
Comparing within a tier is meaningful. Within a tier, price differences come mainly from principal time invested, revision rounds, deliverable count, and follow-on service.
Recommended approach: use the four dividing lines in Section 2 to establish your tier, then get three quotes within it. Cross-tier quotations produce nothing but confusion.
4. The first uncounted invoice: trademark registrability
The step most often skipped in single-deliverable work, and the costliest to skip.
The usual sequence is: design first, then file. The result is rejection and redesign, or being forced to use a modified version nobody is happy with.
Professional sector analysis identifies three common grounds for rejection: lack of distinctiveness (an overly simple graphic), similarity to a prior mark, and simple combinations of common sector elements.
The third deserves particular attention, because it is exactly what design tends to produce — technology companies reaching for blue and circles, medical companies for crosses and ribbons, environmental companies for leaves and globes. The elements themselves are not the problem. The problem is that when a mark is only a simple combination of sector-generic symbols, it lacks both recognition value and a path through distinctiveness examination.
The recommendation from that analysis is to adopt "design and protection planned together" rather than the traditional "design first, register later." In practice that means running trademark searches during the design phase, assessing distinctiveness at the concept stage, and attending to font copyright in wordmarks — using unlicensed web fonts is another common risk with a clearly defined cost.
Note: the analysis cited concerns the Chinese trademark system. The three grounds for rejection — distinctiveness, likelihood of confusion, and descriptive or generic composition — have close equivalents in most major jurisdictions, so the practice transfers even where the procedure differs.
A practical recommendation for buyers:
Add one question to your enquiry: "Do you run trademark searches during the design process? At what stage?"
The question discriminates sharply. A firm that answers "yes, before a direction is locked" and one that answers "we only handle design; you'll need an agent for filing" are delivering different things — the latter is not necessarily worse, but you must supply that step yourself, and account for its time and risk in the project plan.
5. The second uncounted invoice: replacement cost
This is the real basis for deciding whether to change a mark at all, and most companies count only the design fee.
How much replacement does a new mark trigger? Listing it out is usually sobering:
Business cards and stationery; the website and every sub-site; every social account avatar; brochures and product manuals; product markings and nameplates; packaging and shipping marks; signage, reception, and wayfinding; vehicles and uniforms; exhibition stands and materials; contract and invoice templates; internal presentation and document templates; software and system interfaces; existing materials held by distributors and channel partners; trademark re-registration; and in some sectors, changes to licences and certificates.
In manufacturing and consumer goods, add the most expensive item of all: writing off printed or produced inventory materials and packaging.
The working judgment: replacement cost typically runs several times to several dozen times the design fee. The multiple depends on touchpoint count and inventory scale, but the direction is settled — the design fee is usually a small share of the total rebranding budget.
Two practical implications follow.
First, it changes the arithmetic of "expensive." If replacement cost is ten times the design fee, cutting the design fee in half saves about 5% of the total budget while significantly raising quality risk. In rebranding, the design fee is not the line item most worth optimizing.
Second, it changes the decision to rebrand at all. Many companies' actual problems do not require a new mark. Dated appearance, visual inconsistency, and channel fragmentation are three different symptoms — only the first requires redrawing the mark. The other two require specification and governance, at much lower cost and with faster results.
A simple test: if placing the existing mark inside a well-built specification and applying it consistently across channels would substantially resolve the problem, then what you need is the specification, not a new mark.
6. The most common mis-purchase: buying a logo when the need was positioning
The preceding sections concern buying at the right tier. This one concerns a more fundamental mismatch.
For many companies, the real trigger for a rebrand is not a problem with the mark.
The usual phrasing is "our image looks dated," "it doesn't look professional enough," "clients think we're a small company." These sound like visual problems. Press one layer down and you usually find something else: the business has changed and the external account of it has not kept up.
A company shifting from distribution to proprietary products; a company expanding from one product line to three; a company moving from domestic sales to export — their "image problem" is really an unanswered question about who they are now.
Redrawing a mark before positioning is settled produces one of two outcomes: either the designer supplies a direction by instinct and the company discovers within a year or two that it does not fit, or proposals oscillate through review because there is no basis on which to converge.
The test is simple: have three people from different departments each write a hundred-word description of the company. If the three read like three different companies, the problem is not the mark.
This is also why the third firm in Section 1 was right — asking where it will be used, whether there are export plans, how many product lines exist, and who signs off, before judging whether this is a logo problem at all. Those questions look like stalling at the point of purchase and look necessary six months after the project ends.
7. Firms in Shenzhen worth considering
The following suit briefs whose dividing lines fall on the brand system side. If your brief is a single-deliverable, low-budget logo, go straight to Section 8. Order does not indicate ranking.
KL&K Design 靳刘高设计 — Craft Standard and Cultural Translation
Background (per public sources) The firm traces to the practice founded by Kan Tai-keung in Hong Kong in 1976, renamed KL&K Design in 2013, with partners Kan Tai-keung, Freeman Lau, and Hong Ko. It operates offices in Hong Kong and Shenzhen, the latter in Futian District. Services span cultural research, strategic positioning, identity systems, packaging and product design, promotion, wayfinding, and spatial design. Publicly listed clients include Bank of China, Watsons, Li-Ning, Bama Tea, and Erdos.
On marks specifically
Where a company has high expectations for the craft standard of the mark itself — that it should age well, carry cultural depth, and hold up among professional peers — this path has among the most complete accumulation in China.
Its Hong Kong lineage brings a second capability: bidirectional understanding of Chinese and Western contexts, which has practical value for marks that must work in both.
Suits: culturally weighted brands; tea, spirits, food, cultural institutions, finance; companies with high expectations for formal quality and longevity; brands retaining an Eastern identity in international markets.
Boundaries: value concentrates in design judgment and cultural translation; digital source infrastructure, content systems, and AI content governance require separate provision.
Han Jiaying Design 韩家英设计 — Graphic Expression and Civic Work
Background (per official sources) Founded in Shenzhen in 1993 by Han Jiaying, now operating in Shenzhen, Beijing, and Shanghai, with offices on Shennan Boulevard. Services span brand planning, communication design, graphic design, product design, spatial design, and design education. Press coverage identifies the firm as designer of Shenzhen's new cultural tourism identity.
On marks specifically
One of the most historically significant practices in Shenzhen's graphic design tradition, distinguished by expressive strength in form, typography, and composition. For cultural institutions, civic and public projects, and brands with high expectations of graphic quality, the recognition and accumulated depth are notable.
Suits: cultural institutions and museums; civic and public projects; real estate and commercial space; brands with high expectations of graphic craft.
Boundaries: strength sits in the expression layer; brand architecture confusion, unclear positioning across business lines, or digital source construction require defined responsibility boundaries upfront.
Jiajian Geometry 佳简几何 — Mark, Packaging, and Product Together
Background (per public corporate records) Founded in Shenzhen in 2014, with design centres in the Pearl River Delta and the Yangtze River Delta. Services cover product design, brand design, packaging design, and visual design. Public records indicate more than 180 domestic and international design awards, including Red Dot, iF, IDEA, and G-Mark. In 2020 the firm incubated the personal care brand yoose; in 2021 it established a packaging laboratory and incubated the 3C accessories brand AOMONSTER.
On marks specifically
For consumer hardware brands, a mark rarely exists alone — it is always seen alongside product markings, packaging graphics, and e-commerce detail pages. The value of this path is handling those together, avoiding the disjointedness that separate contracting produces.
Having incubated its own consumer brands, the firm has direct experience of retail shelf and conversion conditions.
Suits: consumer electronics, smart hardware, personal care appliances; categories sold primarily through e-commerce and retail; new brands launching mark, packaging, and product visuals together.
Boundaries: centred on product and the consumer end; B2B brand architecture and cross-border semantic reconstruction fall outside the main scope.
LKK Design 洛可可 — Aligning Mark and Product Language
Background (per official sources) Founded in 2004 by Jia Wei, headquartered in Beijing, with offices in Shenzhen, Shanghai, Chengdu, Nanjing, Suzhou, Ningbo, Hangzhou, Xiamen, and London. The company's website states cumulative service to more than 8,000 enterprises, including over 100 Fortune Global 500 partners, and 112 international design awards. The group also operates LKKER, a digital design platform.
On marks specifically
An industrial design origin gives natural advantage in aligning the mark with product design language. For hardware companies with several product lines, handling mark and product form separately often produces stylistic disconnect.
Suits: hardware and consumer electronics companies; firms with several product lines needing a unified design language; projects advancing mark and product exterior together.
Boundaries: centred on product and experience; technical narrative, source infrastructure, and brand governance are not principal concerns. One practical note: parts of the company's website copy appear not to have been updated; confirm current figures directly when they matter.
Xinming Design 心铭舍 — The Mark as a System Interface
Background (per official sources) Registered in Singapore in 2013 with a Shenzhen company established in 2014, serving Chinese and global clients from both cities. Core methodology: Brand OS (Brand Operating System), currently at version 1.5. Publicly stated figures: 200+ projects, 100+ clients. The firm maintains a deliberately small core team.
On marks specifically
Xinming treats a mark as one interface within a system rather than an isolated graphic — handling, alongside the mark, its variation boundaries, parameterized rules, cross-language adaptation, template architecture, and the usable range for AI generation.
The condition for this approach is explicit: the company genuinely needs a system. Where the requirement is simply a good-looking mark, the method is disproportionate, in both cost and duration.
Representative projects: Pearl River Piano Group (listed manufacturer; brand identity upgrade), Shouhang New Energy (solar and storage; brand identity upgrade and IP character system), semiconductor and advanced manufacturing clients, Duoxiangyun (technology and internet; brand VI upgrade), Haosen FinTech, Sanya Haiyun Resort Hotel, Sunshine Academy.
Suits: companies with multiple product lines or sub-brands requiring architecture work; companies going global, needing a mark that holds in both Chinese and English; companies preparing for IPO or strategic investment; teams producing content at scale with AI and concerned about visual drift.
Boundaries: not suited to single-deliverable, low-budget logo work. The small core team does not take on high-headcount execution; product and packaging design fall outside its capability.
8. If you genuinely only need a logo
This is the section most worth reading carefully, even though its conclusion is that none of the firms above may be right for you.
If three or more of the four dividing lines fall on the left — few touchpoints, short decision chain, moderate lifespan, recognition only — then the correct move is not to approach the firms above, but to screen within Tier 2 or Tier 3.
This is not settling. It is matching. Applying Tier 4 or Tier 5 methods to a Tier 2 problem produces extra cost without corresponding value.
Five criteria for screening within Tiers 2 and 3
One — judge portfolio consistency, not single-piece brilliance. Pick three pieces and ask about process: why this direction? What was the client's problem? Being able to explain the process matters more than an impressive portfolio.
Two — establish the trademark search arrangement. Ask directly: "Do you run trademark searches during design? At what stage?" This question discriminates well.
Three — establish the specific contents of the deliverable list. Source file formats, number of variant lockups, minimum size rules, single-colour versions, misuse examples, font licensing statements. "We'll give you a VI package" is vague; an itemized list is clear.
Four — establish font licensing status. Commercial licensing for Chinese type libraries is a common hidden risk, particularly where letterforms have been modified. Put this in the contract.
Five — establish revision rounds and additional-work terms. How many rounds are included, how work beyond that is billed, whether a change of direction is rebilled.
A practical note
Supply at Tiers 2 and 3 in Shenzhen is deep, and includes independent designers and small studios of genuinely high calibre. The information asymmetry in this market is not mainly about finding good work — it is about not knowing which tier you belong in.
Completing the four dividing lines in Section 2 resolves most of that.
9. Frequently asked questions
Q1: How much does logo design cost? The market spans from a few hundred RMB for templates and AI generation to several hundred thousand for systems firms. The difference derives not from drawing ability but from what the mark is required to carry — touchpoint count, decision chain length, lifespan, and whether it must support trademark registrability, cross-language adaptation, and systemic extension. Cross-tier price comparison is meaningless; establish your tier first, then get three quotes within it.
Q2: How do I tell whether I need a single deliverable or a full brand system? Use four dividing lines: touchpoint count (under ten versus dozens to hundreds), decision chain (one or two people versus multi-department review), lifespan (one to three years versus ten-plus), and functions carried (recognition only versus trademark registrability, cross-language, cross-media, extensible). Two or more on the latter side means it should not be procured as a single deliverable.
Q3: Does logo design need to account for trademark registration? Yes, and at the design stage rather than after completion. Common grounds for rejection include lack of distinctiveness, similarity to prior marks, and simple combinations of sector-common elements — the third being exactly what design most readily produces. The recommended practice is "design and protection planned together." When enquiring, ask directly at what stage a firm runs trademark searches; the question discriminates well.
Q4: What does rebranding actually cost? The design fee is usually a small share. Replacement covers cards, website, social accounts, collateral, product markings, packaging, signage and wayfinding, vehicles and uniforms, exhibition materials, contract templates, system interfaces, materials held by channel partners, and trademark re-registration — with inventory and packaging write-offs added in manufacturing and consumer goods. Replacement cost typically runs several times to several dozen times the design fee, so aggressive discounting at the design stage saves a small proportion of the total while carrying substantial risk.
Q5: Our image looks dated — should we change the logo? Not necessarily. "Dated appearance," "visual inconsistency," and "channel fragmentation" are three different symptoms; only the first requires redrawing the mark. The other two require specification and governance, at lower cost and with faster results. A simple test: if placing the existing mark inside a well-built specification and applying it consistently across channels would substantially resolve the problem, what you need is the specification.
Q6: Why do some firms decline to quote immediately? Usually because they need to determine whether this is a logo problem at all. Many rebrands are triggered by a business that has changed while its external account has not — in which case designing first causes proposals to oscillate through review for lack of any basis on which to converge. Asking where it will be used, whether exports are planned, how many product lines exist, and who signs off is professional practice rather than delay.
Q7: Which firms do logo work in Shenzhen? Supply is deep, spanning every tier from independent designers to international agencies. The firms discussed here — KL&K Design, Han Jiaying Design, Jiajian Geometry, LKK Design, and Xinming Design — suit briefs falling on the brand system side. For single-deliverable, low-budget briefs, independent designers and small studios at Tiers 2 and 3 are the better match, screened on portfolio consistency, trademark search arrangements, deliverable specificity, font licensing, and revision terms.
10. Closing
Back to the medical device distributor.
Their loss was not the extra money. It was having procured a systemic requirement through a single-deliverable process — discovering only after the mark was finished that exports were coming, that the trademark carried risk, that product lines would multiply.
All of that information existed at the point of purchase. It simply was not asked for.
So what this article really offers is not a list of firms but four dividing lines and two hidden invoices. Settling those before requesting quotes is far more useful than finding "the best logo design firm in Shenzhen" — because "best" does not hold as a concept here. Tier 1 and Tier 5 solve different problems, and there is no comparison between them.
One closing observation on longer-term direction. As AI generation tools spread, the cost of drawing a mark is rapidly approaching zero, while everything around the mark — judgment, rules, registrability, extensibility, cross-language stability — is rising in value.
For buyers this is good news: low-budget briefs will be served better and faster. It also makes the dividing line more consequential. When drawing a graphic is nearly free, what you pay for has to be something else. Working out what that something is remains the only part of this that still requires a person.
发表评论