Cat Carrier Pet Canton Fair: Trade Event Operations

Pet carrier production desk · Updated 2026-10-06 · 15 min read

Most buyers ask a pet bag supplier the same opening question: can the cat carrier line be reordered in the original colour six months later? The answer depends on dye-lot control, not on goodwill.

Conversation Volume: Capture and Triage at Scale

The defining number at a large general fair is not the size of the booth but the number of conversations per day, and it is high: 180-420 on a busy day against 18-46 at a specialist show. At that volume, conversation quality is worthless without a capture system, because no one remembers 300 people.

The failure mode is well known and still common: a team collects 1,200 business cards over five days and qualifies none of them, because no record exists of what was discussed. A booth with a structured capture process and half the card count produces more orders, and the difference is entirely process.

The mechanism is a graded card. A two-second grade assigned at the moment of the conversation, A for specification and quantity mentioned, B for product interest without quantity, C for a general enquiry, means the follow-up can be sequenced by quality rather than by chronology. Grading costs nothing and it is the single highest-value process at the fair.

The arithmetic that follows is the argument for grading. Over five days, a booth collects roughly 55-250 A-grade conversations converting at 18-34%, which is 10-85 orders, against 305-1,010 C-grade converting at 0.6-2.4%, which is 2-24 orders from ten times the follow-up work. Following up everything equally means spending most of the effort on the smaller return.

Digital capture beats paper at this volume. A tablet form with twelve fields, filled in while the buyer is still at the booth, produces a record that can be acted on the same evening, and it eliminates the transcription step that loses 8-22% of paper cards. Setting up the form costs 200-900 USD and it handles 1,200 conversations without additional effort.

Photographing the buyer with the samples they discussed is the trick that makes a follow-up specific. A photograph attached to the record costs five seconds and lets the day-one email reference the exact product, which lifts response from 38-62% to 52-78%.

Grade every conversation A to D at the moment it happens: A-grade converts at 18-34% against 0.6-2.4% for C, and following up everything equally spends most of the effort on the smaller return.

Sample Strategy for a General Buyer Pool

A general fair brings buyers from every category and every market, many of whom have no prior knowledge of pet carriers. The sample set has to work for a specialist buyer who wants seam construction detail and for a general buyer who wants to understand the category at all.

The answer is breadth with two depths. Breadth covers every size and the main construction types so that any buyer can locate their interest. Depth is concentrated on the two or three products the programme actually wants to sell, which carry cut-aways, hardware options and finish variants.

The count is higher than at a specialist show: 46-88 units against 28-46, weighing 104-224 kg and occupying 2.4-5.2 m3. That is a real freight and handling cost, and it is the price of a fair where the buyer pool is undefined.

Trim and hardware boards are cheap and disproportionately useful at a general fair. A board carrying four zipper gauges, three buckle types, four webbing widths and the available finishes lets a buyer specify without a sample being made, at 40-180 USD to assemble and 2-6 kg to ship. It converts a vague conversation into a specification in about three minutes.

Durability of samples matters more at this scale. A unit handled 150-400 times a day across five days will show wear, and a visibly worn sample on day four undermines the product. Rotating the display stock, with replacements brought out each morning from locked storage, keeps the floor set presentable and costs nothing but discipline.

Theft and loss is a real line at a general fair. Expect 1-4% of a 46-88 unit set to go missing over five days, which is 1-4 units, and treat it as a cost of the channel rather than a failure. Keeping the depth products on a tether and the breadth products free to handle balances the two.

Bring 46-88 units for breadth with depth on two or three products, plus a trim and hardware board at 40-180 USD that turns a vague conversation into a specification in three minutes.

Quotation Turnaround: Hours, Not Days

At a general fair a buyer meets twenty suppliers in a day and expects a number the same evening. A quotation delivered four days later arrives after the shortlist is closed, and the fair's value is lost regardless of how good the quotation is.

The target is 4-18 hours from conversation to a quotation in the buyer's inbox. Achieving that at 180-420 conversations a day requires pre-built quotation templates rather than individual work: a specification sheet per product, a price ladder per product, and an adder table, assembled into a document in 10-25 minutes per enquiry.

Templates have to be market-aware. A buyer from the United States, one from the European Union and one from Japan need different compliance references in the quotation, and pre-building three variants costs 200-900 USD and removes the per-enquiry research step.

The last row is the warning. An enquiry answered after the fair closes converts at 1.4-4.2%, against 16-30% for the same enquiry answered the same evening. The difference is not the quality of the answer; it is that the buyer has already shortlisted someone else by then.

The evening production cycle is the operational answer. Running a quotation shift from 18:00 to 23:00 during the fair, with two people and the capture records from the day, produces 20-60 quotations a night and clears the A-grade backlog before the next morning. At 40-120 USD of staff cost per night, that is the best-spent money at the fair.

What must never be sacrificed for speed is the assumption statement. A quotation that quotes 15.40-16.90 USD at 500 units without stating the material set, the construction, the payment terms and the port is a quotation that will be renegotiated later. Stating the assumptions costs two minutes and protects the price.

Answer within 4-18 hours using pre-built templates: the same enquiry converts at 16-30% same-day against 1.4-4.2% after the fair closes.

The Visit That Follows: Hosting a Production Base Audit

At a large general fair the floor conversation is a filter and the production base visit is the decision. Expect 22-38% of A-grade buyers to request a visit, and expect it to be requested within 3-9 days of the fair closing. That visit, not the booth, is where the order is won.

A visit is a three to six hour event with a predictable shape: a meeting, a line tour, a quality station, a sample room, and a closing commercial conversation. Each part answers a question the buyer could not answer at the booth, and each can be prepared in advance.

The line tour is the part buyers remember. What they are actually assessing is order and control: whether material is identified and segregated, whether work in progress is tracked, whether operators have work instructions at the station, and whether defects are captured rather than hidden. None of that requires new equipment; it requires that the existing system is visible.

The sample room is underrated and it is where a visit often turns. A buyer who came for one product and sees a coherent range, development samples and evidence of engineering capability leaves with a larger programme. Maintaining a sample room costs 240-1,400 USD to set out properly and it is the cheapest capacity expansion available.

The quality station is where a visit is lost. A buyer asking to see the inspection records for the last three shipments, the AQL sampling plan actually used, and the calibration status of the test equipment is running a standard check, and a programme that cannot produce it in five minutes fails. Keeping those three items in one folder costs nothing.

Logistics matter and they are unglamorous: transport from the fair city, a room that seats six, lunch, and a schedule that finishes in time for the buyer's flight. A visit that runs past the buyer's departure time ends badly regardless of content. Booking a return movement with 90 minutes of slack costs 60-240 USD and prevents the most common ending failure.

The visit is the decision event: 22-38% of A-grade buyers request one within 3-9 days, and the three items that must be producible in five minutes are inspection records, the sampling plan and equipment calibration.

Slot Reservation During the Fair Season

A large fair concentrates a year of enquiries into five days, and the production capacity behind it does not expand to match. The result is a queue, and the queue is where a good fair becomes a bad year.

The mechanics are simple. A fair closing in late spring generates orders that all want production in the same 35-50 day window, and capacity that runs at 84-92% utilisation in a normal month cannot absorb a 30-60% spike. Lead times quoted at the fair of 35-50 days become 60-85 days by the second week if nothing is done.

The answer is to reserve capacity before the fair rather than after it. Booking 20-40% of the following quarter against forecast demand, at 300-900 USD per reservation or against forecast commitment, means the fair's orders land in a slot that exists rather than in a queue that forms.

The last row adds the material bank, and it is the more powerful of the two levers. Reserving line time without reserving material produces a slot that starts on the line but waits at the mill, so the two have to be booked together. A material bank holding shell fabric, mesh, hardware and zipper chain against a forecast costs carrying cost at 0.22-0.44 USD per unit per month and removes 14-35 days.

Pricing the slot honestly is the other discipline. A buyer offered a confirmed start date at the standard price is being given something worth having, and a buyer offered an estimate at a discount is being given a problem. Quoting a confirmed date at 35-50 days and an unconfirmed date at 55-80 days, both at the same price, lets the buyer choose and costs nothing.

Expedited production is the release valve and it should be priced and offered rather than improvised. Compressing 35-50 days to 18-26 days costs 1.80-4.20 USD per unit in overtime, air freight on materials and schedule disruption, and offering it as a named option at a named price prevents the far worse outcome of promising it free to close an order.

Reserve 25-30% of the following quarter before the fair and book the material bank with it: quoted 35-50 days becomes actual 28-42 days instead of 62-88.

Compliance Documentation for a Mixed Buyer Pool

A general fair brings buyers from every market, and each market has its own documentation expectation. A single compliance file does not serve them, but twenty files are impractical, so the answer is a modular pack assembled per enquiry.

The core module is common to all buyers: the social compliance audit, the quality management certificate, and a general product safety declaration. Market modules then attach: CPSIA and Proposition 65 for the United States, REACH and general product safety documentation for the European Union, and specific labelling requirements for Japan and Korea.

The most common failure is not absence but mismatch: a buyer from California asking about Proposition 65 and receiving a REACH declaration, which is irrelevant to them and signals that the supplier did not listen. Matching the module to the stated market costs nothing and prevents that signal.

The visit column is the striking one: 78-94% of buyers ask for compliance evidence at the production base visit against 44-68% at the booth. That is because the booth conversation is about product and the visit is about capability, and documentation is capability evidence. Preparing the pack for the visit rather than for the booth is the correct priority.

Test reports should be matched to the samples shown. A buyer examining a specific construction will ask whether that construction has been tested, and a report for a different product is worse than an honest statement of what has and has not been tested. Where a new construction is shown untested, giving the test schedule is more credible than a claim.

Physical testing is referenced to ASTM International methods, chemical declarations are issued against OEKO-TEX criteria, and US market claims are held against Consumer Product Safety Commission guidance. Our production team keeps the modular pack current and matched to the sample set, which is what lets a visit answer a documentation question in minutes rather than days.

Build a core pack plus market modules and match them to the stated market: 78-94% of buyers ask for compliance evidence at the visit against 44-68% at the booth.

The Sampling Queue: Capacity, Sequencing and Wait Times

A large fair generates sample requests faster than any sampling operation can absorb, and the queue that forms is where the fair's value leaks away. Managing it is a scheduling problem with a known capacity and a known demand.

Sampling capacity is finite and measurable: a development team running parallel cycles can produce 8-22 samples in a 6-10 working day window, depending on whether a new pattern is needed. A fair generating 40-120 sample requests creates a backlog of three to eight cycles, which is 18-80 days of waiting.

The consequence is stark. A buyer expecting a sample in 10 days and told 60 has usually moved on, and observed conversion from a sample request falls from 38-62% at a 10-day wait to 6-16% at a 60-day wait. Capping the number of samples offered and sequencing them is better than accepting all requests and serving none well.

The last row is the structural answer and it is the same lesson that appears throughout this programme: a released platform library converts sampling from development work into adaptation work. Where a request matches an existing platform, the sample is a colourway or a branding adaptation at 1-6 days rather than a new build at 6-10 working days, and capacity effectively doubles.

Sequencing by grade is the operational answer within a fixed capacity. A-grade requests with specification and quantity go into the first cycle; B-grade into the second; C-grade receive a sample offer without a commitment. Communicating the schedule honestly, including the wait, is better than a promise that slips, because a buyer told 24 days and receiving it in 24 is still a buyer.

Temporary capacity expansion is possible and priced. Adding a second development shift for two cycles costs 2,400-7,800 USD and raises capacity by 60-90%, which on a fair generating 80 requests is the difference between 24-70 days of wait and 18-40. Against orders worth 7,700-33,800 USD each, that is a straightforward calculation.

Capacity of 12-22 per cycle against 40-120 requests means a 24-100 day wait: sequence by grade, communicate the wait, and use the platform library to turn development into adaptation.

Cost Model: Large General Fair Against Specialist Show

The two fair types can be compared on the same annual budget, and they produce different things. A large general fair produces volume at lower average value; a specialist show produces fewer, larger programmes at higher cost per conversation.

On 30,000-60,000 USD of annual budget, a general fair buys a 9-18 m2 booth for two fairs a year, generating 1,600-4,800 conversations and 10-85 converted programmes. The same budget at specialist shows buys one or two booths generating 180-620 conversations and 2-13 programmes.

The difference that matters is not the count but the average programme size and the cost of the conversations that did not convert. A general fair's C and D grade conversations consume enormous follow-up capacity for a small return, and that cost is real even though it does not appear on an invoice.

The general fair wins on cost per programme decisively, at 353-6,000 USD against 1,385-32,000 USD, and that is why it remains the largest item in most export programmes. The specialist show wins on programme size and on the quality of the visitor, which is worth more than the cost comparison suggests for a supplier selling a considered, specification-driven product.

The correct answer for most programmes is both, sequenced by maturity. A programme without a track record should run the general fair to build volume and learn the buyer pool, and add a specialist show once it has a platform library and a compliance pack worth showing. Running the specialist show first, with a thin range and no documentation, wastes its advantage.

Commercial terms are identical at both: MOQ 500 pieces per colourway, prototypes in 6-10 working days, bulk production 35-50 days after sample approval, final random inspection to AQL 2.5, T/T 30/70 and FOB Xiamen. Our production team quotes from the same ladder at a general fair as at a specialist show and at a production base visit, with process documentation under ISO 9001 coverage. The general fair costs 353-6,000 USD per converted programme against 1,385-32,000 USD for a specialist show; run the general fair to build volume and add the specialist show once the platform library and compliance pack exist.