Hi Sasha, and welcome to the community!
I’m an integration specialist who typically works on very difficult projects so I’m not here to pitch services. Rather, I’m here to offer a few comments about your integration strategy and to share some tips that the community might find helpful.
Airpower has a lot of good reviews, but not because of its price. :winking_face: It seems expensive and it is depending on your store’s volume. But even with this price, it may be a bargain. Integrations that update/synch an inventory looks pretty simple on the whiteboard, but it’s actually a complex beast in some situations and depending on your integration requirements which exposes good cause for tip #1.
Tip #1 - Process Requirements
Always create detailed requirements in advance of actually building the solution. If you ask your chosen consultant to build it without clear requirements, you may endure rework, rewrites, and some agony as you try to roll out that perfect solution that allows you to get your weekends back (if ya’ know what I mean).
It’s often a good idea (depending on the scope) to separate the goals of the process from the implementation tasks primarily because the writing out goals deeply influence the actual implementation tasks.
Tip #2 - The Definition of Sync
The obvious isn’t [typically] so obvious to consultants and your definition of “syncing” may be completely different from your consultant’s definition. Make sure you both agree at a fairly detailed level that the process will meet your expectations. Depending on the integration consultant, you might benefit from literally writing out what it means for a successful synching to occur. Indeed, the best approach is to craft a number of typical inventory scenarios - examples, …
- How does the integration process handle returns?
- How would the system deal with partial shipments and backorders?
- How does the process deal with network failures? Does it self-heal?
Tip #3 - Sustaining the Solution
This is the part of the project that is often the furthest from your mind and almost certainly not a consideration for integration developers because they’re hired guns - they resolve [today’s] challenge and rarely build something that anticipates future changes. Most integrations fail within 30 days of initial deployment. Many approaches become a collection of patches and are often abandoned for a new component or idea.
The’s one thing that we can all agree on concerning glu-factories like Zapier - they fail often and they degrade quickly when anything changes. You may believe your database design is in stasis, but no database is without regular updates to accommodate new innovation, new customer needs, and new business requirements from compliance to security, and all things between.
Plan on change; encourage your consultants to select tools that are sustainable. A good example - while Zapier provides a way to fully automate sync processes without code, they also offer access to a code framework. But this is an afterthought for Zapier - it was built by popular demand, not as an underpinning for the code-free movement. Consider alternatives like Autocode which was developed as a stringent NodeJS platform first; then it began to offer prefab recipes all based on actual NodeJS underpinnings. This makes it far more likely to support not only very complex integration requirements but also simpler recipes that can be easily maintained.
There are more than 70 published integration recipes for Shopify at Autocode and almost all of them created since mid-2020. And the free tier is pretty generous (one million API requests per month).
ps - I don’t work for Autocode and my links are not bound to an affiliate promotion - I just have admiration for the approach those developers have towards integration processes.
