Three months ago, a developer experienced something like this: the search function they had created for their project completely stopped working after being totally reliable for several months. The developer had had no problems returning results from their custom search until about three weeks ago, when they logged into the account and found that they were receiving the following error: “The custom search cannot be returned as this application is not authorized to make use of the JSON API for Custom Searches.” Essentially, what the error message is saying is that both the application and keys used to authenticate the application remained the same at the time of initial development.

When this issue occurred, there was more involved than what might seem like an issue to one developer; there were also concerns that the number of third parties that developed search technology based on API’s were having issues due to the way in which these technologies were interacting with the API, and Google was taking a larger position on this issue. While no one could say with certainty how the new technology would continue to play a role in relation to developing search capabilities, many were left wondering what the implications for their search systems would be going forward.

Something That “Just Worked” Suddenly Isn’t Working

Many developers use the tools provided by Google Programmable Search extensively; in fact, many developers have become very dependent upon them and rely on Google Programmable Search to do much of their searching. These were inexpensive, easy to integrate with other services, and widely used. Industry estimates indicated that thousands of smaller and mid-sized applications relied on these APIs for functionality needed for both their internal dashboards and their external searching capabilities.

When these tools started generating unexpected errors, there was much confusion among developers who described the problem in virtually identical terms on the forums. Each developed provided no configuration changes or billing issues, but found themselves unable to access their service at the time they were trying to use it (403 errors). A 403 error (in context) is much more than just a small bug; it is an indication that the service itself may not recognize or support the end user’s project anymore.

As one developer stated, “It worked yesterday; today it doesn’t; and nothing has changed.” This simple statement sums up the real issue at stake here, which is reliability.

The Importance of Search as the Backbone of User Experience

Search is more than a feature; it’s the anchor of the user experience with your product. Over 43% of users will use the search feature on the first page of their visit to any given website, while nearly 70% of e-commerce sales are indirectly driven by users utilizing your site’s search function in some fashion.

When a user can’t find what they’re looking for through search, they will leave your site. That’s why when there are unanticipated issues with third-party APIs being down – even for a few hours – it can result in severe loss of sales, user confidence, and your ability to operate efficiently. More importantly, it identifies a problem that many startups underestimate: relying on external platforms that they don’t control.

Looking for a reliable alternative to avoid search downtime?

AI-powered site search built for reliability and control

Never let search failures impact your users again

ExpertRec helps you build a stable, customizable search experience without relying on unpredictable third-party APIs. Get full control, better performance, and zero downtime risks.

Start Free Trial →

Quick setup • No coding required • Works on any website

Significant transformation of search infrastructure.

Over the last few years, search technologies have changed dramatically without receiving much notice. The implementation of newer systems is happening with a greater level of control, personalized features, and greater emphasis on user privacy. Some key drivers for this shift are cost reduction, improved performance, and an increased focus on independence. Recent estimates state that 60% of businesses are now looking for substitutes to their current use of third party API’s to perform critical functions such as searches. The overall purpose of moving to a more “predictable” search experience is the reduction of unplanned outages caused by not having ownership or control over the search experience.

The Role of Expertrec

Expertrec was created to help developers build custom search applications with complete sensibility. The no-tolerance attitude toward using third-party APIs means that any external API has the potential to change without warning. Expertrec’s model enables developers to build their own search application, using Expertrec technology, independent of third-party organizations whose service offerings could change overnight.

With this model, you do not have to be a computer science major or have technical knowledge to set up a search management system for your company that can quickly respond to the rapid growth of both new customers and new products that are being developed for a growing market.

As stated above, having the ability to control your search application’s evolution is likely the greatest single benefit to any developer building their custom search solution. Because your search application is being developed to integrate directly with the back-end as well as with your internal processes, you are able to control how your search application continues to grow and evolve. This also means that you will NOT have to rely upon third-party vendors to keep your search solutions aligned with your goals and your progress as you work toward bringing your company’s vision into reality.

What Should You Do Next?

When your search API fails today, your first step is to check with your provider for access rights and billing issues (may be only temporary). If it still does not work or if it does not seem reliable, you should begin exploring more reliable alternatives.

Fixing an issue today does not help you tomorrow; you are losing potential product development dollars if you are not planning for future failures.

If this experience teaches you anything, it may be how quickly the systems relied upon may change.

Final Thoughts

Your users shouldn’t have to think about searching; it should be seamless and instantaneous, reliable. The second it doesn’t work properly, it becomes the most glaring defect in your product.

To give an example of the current situation, relying solely on a third-party provider for the functionality you require is risky; it’s worth exploring other options like Expertrec or others; it’s not just a matter of you fixing a problem, but making a strategic decision.

Also, the best time to think about your search application is not when it fails, but before it fails.