winding road

When Microsoft made the push to the cloud it hit a wall of resistance. 

Many organizations hesitated to go all-in on the cloud, opting instead for a hybrid environment, spreading their content and applications across on-premises and the cloud.

And sooner or later, it became unavoidable for these same organizations to also implement hybrid search. 

What's Under the Hood in Hybrid Search

Microsoft announced a new hybrid search model at its inaugural Ignite Conference in 2015. Hybrid search allows businesses to index both on-premises content as well as cloud content into the cloud index. This new paradigm provides the benefits of aggregating results from on-premises and the cloud in a single, unified result set, which benefits from an integration with Office Graph as well.

Office Graph is the most significant representation of Microsoft’s commitment to making machine learning techniques available for everyone. Office Graph tracks what Microsoft calls "signals" and uses those signals to understand the relationship between users and the interactions between users and information. 

Office Graph and Delve form the basis of Microsoft cloud-based discovery application. Delve is a user-facing application in Office 365 that helps users discover relevant content from across the organization’s tenancy. However, by using hybrid search, Delve can also display on-premises content, too.

From a maintenance perspective, it’s important to note that by using a hybrid search approach, the main content processing and indexing components are in the cloud. This significantly reduces the need for on-premises resources.

6 Common Hybrid Search Mistakes (and the Solutions)

These benefits led more and more companies to plan and implement hybrid search. However, there are several common mistakes which are better to avoid from the very beginning:

1. Ignoring hybrid search

Some organizations don’t see the value in hybrid search, rather choosing to keep separate search applications for both on-premises and online. The primary reason is that they don’t see the benefits, therefore they won't commit to investing in it.

How to avoid it: Organizations with content both in the cloud and on-premises should at the least consider how a hybrid search solution can help to improve your employees’ everyday performance. A Proof of Concept (POC) is also a good idea, to present these benefits, without any significant investment or long-term commitment.

2. Dive “all in” immediately

In some cases, when the organization sees the clear benefits of a good hybrid search application, it gets so excited it wants to add all content to their search immediately. 

How to avoid it: While the intent to include as much valuable content as possible is commendable, the best practice is to start small and extend search in phases.

3. Looking for one search that “rules them all”

The debate about one centralized search center versus smaller, targeted search applications is ongoing. And while there’s no definite answer, failing to consider search-driven applications can be a big missed opportunity for organizations implementing search.

How to avoid it: While the need for a centralized enterprise search center is unquestionable, always discuss the actual needs and requirements to identify where a specialized, targeted search application can be a better fit. These applications provide much higher value and much better productivity. Your users will thank you.

4. Not considering the benefits beyond aggregated search results

As I mentioned above, hybrid search solutions provide several benefits beyond aggregated search results. Not considering these benefits might lead an organization to the conclusion that the ROI of a hybrid search approach is insufficient to justify investment, even if this results in making a wrong decision.

How to avoid it: Identify all of the benefits for your organization and include these into your ROI calculation.

5. Not involving all necessary stakeholders

Implementing successful search requires several team members with different expertise. If an organization wants to reduce the size of their search team, that quickly results in getting reduced benefits as well, due to the missing expertise.

How to avoid it: Your search team should consist of the proper experts in both your business domain as well as technology. You also need someone who can “translate” between the business and IT, to guarantee that the requirements get implemented the optimal way.

6. Seeing search as a one-time project

Last but not least, businesses should know from the start: there is no definite end for a search project. Long after implementation, the work of governing and managing the search initiative continues — and businesses should plan for such. 

If you do not allocate the proper resources, don’t take care of it, neglect maintenance and updates — your search will get messy very soon. The results lose their relevancy, users do not get the experience they want, and then they stop using it. 

How to avoid it: Your work doesn't finish after you set up and implement search. As the environment changes, so does your business. User needs evolve, and search has to evolve with them to remain successful.

Balancing the Rewards and the Risks

Implementing hybrid search is rewarding, but also risky. 

Each organization needs a unique approach to support their unique requirements. But with proper planning and preparation, you can avoid some common mistakes.

Title image Justin LaBerge