On July 22 this year, the General Office of the Ministry of Industry and Information Technology issued a document, with the document number Gongxinting Qiye Han [2026] No. 353 and a somewhat convoluted title: "Guidelines for Cultivating 'Small, Fast, Lightweight, and Precise' Digital Products and Services (2026 Edition)."
Many people glance at it and scroll away, thinking it is an official document from the authorities setting requirements for service providers and has nothing to do with their own boss.
On the contrary. On the surface, this document regulates the supply side, but in reality it hands the demand side a yardstick. The government is essentially saying: SMEs don’t have much money or enough people, so don’t get burned again by those 'big and comprehensive' projects that cost millions and take two years to implement. What makes a good product? Miniaturized, fast, lightweight, and precise ones are.
We do Odoo implementation in Guangzhou, and our clients are basically small and medium-sized manufacturing and trading enterprises, so we have a deep feeling about this. Let's break down the five yardsticks and talk about them.
The First Yardstick: Miniaturization—Don't Be Greedy for Size
The exact wording in the guidance is “deeply cultivate niche scenarios, accumulate industry knowledge, and control resource usage,” and it also says to avoid piling up “big and comprehensive” features.
In other words: a 50-person injection molding factory does not need a system that covers all the needs of a listed group. What you need is to manage production work orders, procurement, and inventory clearly.
When selecting a product, there is a simple test: have the vendor give a demo. If more than half of the features leave you confused and don't match your business, then it's most likely a generic product being forced to fit, not something made for your industry.
The second yardstick: speed — be wary of projects that show no results within three months
The guidelines require "low-threshold deployment and short-cycle launch," and accompanying training and after-sales support must keep up.
In our experience, for digitalization projects at small and medium-sized enterprises, the best cadence is two to three months per phase, first getting one or two of the most painful processes running smoothly. Those cases where a contract is signed and a team is stationed on-site for half a year without going live are not a technical problem; it's that the requirements haven't converged—or rather, it's that the vendor never intended to let you see results quickly.
The third yardstick: Lightweight — data must be portable
This is the one I consider the most weighty among the five yardsticks. The guidelines explicitly state "support third-party compatibility, support low-threshold secondary development, and ensure convenient data migration", the reason being to avoid "data silos" and "vendor lock-in".
The competent authorities wrote the four characters "vendor lock-in" into a ministerial-level document, which shows just how widespread this problem has become.
What does lock-in mean? It means your customer information, order records, and production data are all in someone else's system. One day when you want to switch vendors, they tell you the data can't be exported, the API costs extra, and the renewal price doubles. Over the years, we've taken on quite a few projects that switched to Odoo midway; the data exported from the previous system was a pile of unusable spreadsheets, and the customer's business history was effectively cut off.
So before signing, make sure to ask clearly about three things: whether the data can be fully exported, in what format, and whether there is an additional fee. If they cannot answer smoothly, be cautious.
The fourth yardstick: precision — results must be measurable
The guidelines require that "results be quantifiable and value perceptible," and also require establishing a closed loop for user feedback.
To put it plainly, digitalization isn't about buying software; it's about making an investment. How many days faster is inventory turnover? How many person-days has reconciliation saved? By how many percentage points has the on-time order delivery rate improved? If these accounts can't be calculated, the project is just buying the boss some psychological comfort.
The fifth yardstick: Digital intelligence—AI is the trend, but don't put the cart before the horse
The guideline encourages products to “deeply integrate artificial intelligence technology,” which aligns with the general direction. But one reminder: AI is premised on data. Your orders are still in Excel, and inventory is kept in the warehouse keeper’s head—no matter how good the algorithm is, it has nothing to work with.
First get the business online and the data structured; only then can AI have a role to play. If this order is reversed, the money spent is basically wasted.
Finally, let's say something practical
The guidelines also specifically mentioned "eliminating hidden charges and bundled sales," and required all localities to build a high-quality resource pool that is "small, fast, lightweight, and precise" to reduce SMEs' selection costs and trial-and-error risks.
For SMEs, the most practical way to read this document is to use it as a negative example for a selection checklist—any product or service that goes against these five directions, no matter how low the quote, should be treated with extra caution.
When it comes to choosing a system, the product is only the starting point—implementation capability is what really matters. With a limited budget, choosing a system with an open architecture that can be enabled on demand, paired with an implementation partner who understands your industry, is more important than chasing brands or feature lists.
If you have any specific questions about selection, feel free to talk with us—sometimes one phone call can help you avoid quite a few pitfalls.
