Templates vs custom:
it's the fit, not the price.

Start with the awkward bit: most H5 projects are fine with a template. Holiday greetings, event notices, a simple brand showcase. It ships on time at a predictable price, with no need to spend tens of thousands more for a vague sense of "premium". Really.

But every year we also inherit projects where the template has run out of road. Rarely a visual problem. It's the one action with no retry: scan-to-enter draws, live sync on site, registration and redemption, screen interaction. Fail on the night and there's no second take.

01Templates get you one. Custom gets you the right one

A template is reuse: one requirement flattened into a fixed structure, development swapped for configuration. Fast, cheap, repeatable. The catch: your idea has to fit the structure.

Custom work is translation. It takes the line in the deck, "make people want to play again", and turns it into interaction, motion pacing, odds design and a performance plan. Slower, dearer, every decision serving this one event.

If it does, go custom. If "having one" is enough, buy the template. The bad ending is paying custom prices for a template, or custom work at template prices. We see both most weeks.

02A template's real bill has four lines, not one

A template quote shows a usage fee. What decides the total sits underneath:

• License scope. Per project, per year or per domain, logo removal extra. Several campaigns across several brands and the licenses stack up.
• Rework. Swapping one image is cheap. Changing the whole visual and motion system isn't. The more complete the template, the more any change drags along.
• Compatibility. In-app WeChat browser, low-end Android, folding screens, each new iOS release. It may never have been tested on all of them, and a white screen after launch means waiting in the vendor's queue.
• On-site cover. Break something on the day and the vendor may be covering dozens of events at once. Where you sit in that queue is a contract question, not an urgency one.

One more cost rarely reaches a budget: the planner quietly cut the most interesting part to fit the template. A discounted idea doesn't throw an error. It just makes the event ordinary.

03Where the money actually goes

Bad news first: a good half of the budget goes to things your deck never mentions. Custom work isn't "more code", it's jobs a template never does. On a mid-complexity H5, effort splits roughly like this:

What each share buys:

1. Brief to plan. Turning "premium", "hype" and "make people share it" into actions and pacing, not guesswork once the visuals land.
2. Design and motion. Type, spacing, transition curves, sound cues. They decide the first impression, and whether a thumb keeps moving.
3. Frontend build. Structure, performance, first-screen strategy. Add a second to load and half the audience never sees the creative.
4. Integration and QA. WeChat engine, low-end Android, share cards, authorized login. Test devices aren't one phone, they're a row.
5. On-site cover. Load testing, fallbacks, hot updates, staff on standby. The event happens once, so the plan is written in advance.
6. Data and review. Tracking, draw odds, anti-fraud, exports. After the event you can still say how many came and where they left.

04When a template is enough. When it isn't.

Run both lists against your project. You don't need all of them. Two or more is enough to start a conversation.

Where a template wins

• Information display dominates, no complex interaction or data capture.
• Brand guidelines loose enough to fit the template's frame.
• Tight budget, tight timeline, standard result acceptable.
• One-off use, nothing to connect to official accounts, mini programs or a member system.

Where custom earns its money

• Real business actions: draws, registration, redemption, check-in.
• Live links to event screens, hardware or third-party systems.
• Brand guidelines the template's structure can't carry.
• The event can't be repeated, and a failure costs more than the build.
• User and behavior data to capture for later.

05An acceptance checklist, ready to use

Template or custom, walk these at acceptance. They rarely show up in a demo, and they're what breaks after launch.

1. The first screen responds within 3 seconds on 4G, no long white screen.
2. Tested on a low-end Android and a mainstream iPhone, scroll and animation included.
3. Share title, description and thumbnail inside WeChat correct, not a stale cache.
4. Draw odds rechecked, winner records exported, anti-fraud rules documented.
5. A fallback page when the network drops or an API fails. Not a blank screen.
6. Tracking covers the key path, raw data exports after the event.
7. Source code, design files, asset copyright and ownership in the contract.

06Picking a template isn't embarrassing. Picking wrong is.

We build custom work, and we keep mature modules. Tools don't rank; fit does. What hurts is mismatched budget and expectation: template money with custom demands, or custom money for a page anyone could have filled in.

Tight budget? Put it into the one action that has to exist. Everything else can be simplified; that action can't be cut. Nobody remembers ten unremarkable pages. They remember one smooth draw, one transition that surprised them, one moment on site that didn't freeze.