Benefits of Coveo Cloud vs On-Premise

I was recently asked by a client about the benefits of Coveo Cloud vs On-premise. So back in 2016 Coveo released of Coveo For Sitecore v4 which coincided with the release of Coveo Cloud v2, this now provided the option of hosting your Coveo for Sitecore index in the cloud as well as on-premise.

Hosting your index in the cloud has some obvious benefits:

  • Speed – developers can be up and running with an index in the cloud in 30 minutes or less.
  • Reduced Infrastructure – you don’t need to procure additional servers for hosting your Coveo indexes across all your various environments.
  • Easier – A Cloud installation of Coveo for Sitecore is easier to maintain than an On-Premises one, as mirrors and redundancy are handled by the Cloud platform.
  • Availability – Cloud architecture is distributed across multiple availability zones and geographical regions for data resiliency and high availability.
  • Scalability – Coveo Cloud solutions are constantly monitored for performance, and measured against performance thresholds and target response times. Built on dynamic processing power environments and thanks to its scalable architecture, Coveo Cloud is able to seamlessly scale computing systems to maintain optimal user experience.

Cloud Only Features

As you can see from Coveo For Sitecore Edition Comparison Coveo Cloud offers features that are only available in the cloud:

  • Machine Learning Relevance Tuning – provides users with a more relevant search result set, based on previous users search interactions. The value is end users can find the content they are looking for much easier and driving them into that sales funnel quicker.
  • Machine Learning Query Suggestion – means users could be presented with alternate searches based upon aggregated, past user behavior. The value here is we get more interaction on the search page from users who might otherwise abandon the search.
  • Machine Learning Recommendations – provide accurate recommendations for related content that reflect what other users making similar searches have also viewed. The value here is you create more opportunities to cross-sell to the user.

Before we take a deeper dive into these features lets first mention the underlying technology behind of these features: Coveo Machine Learning (Coveo ML). This is a cloud and analytics-based machine learning service that continually analyzes search behavior patterns to understand which results and content lead to the best outcomes, such as customer self-service success. In addition to intuitively enhancing search results so the best-performing content always rises to the top, Coveo ML automatically delivers the most relevant search results and proactive recommendations with minimal effort.

Coveo ML continuously learns the evolving user activity and rapidly adapts recommendations following changes such as seasons, new product adoption, or industry news.

CoveoCloudPlatform

Behind the scene, Coveo ML features actually process usage analytics data to build and maintain complex Coveo-managed predictive models to make recommendations.

Now because Coveo ML relies on usage analytics data to start providing relevant recommendations for search results and query suggestions it requires a history search data of at least 3 months.

The Coveo ML Relevance Tuning (or Automatic Relevance Tuning ART) feature analyzes user behavior patterns from many usage analytics search visit actions (such as query reformulation, clicked results, if a support case was submitted) to understand which clicked results and content lead to successful outcomes such as self-service success, and automatically adjusts future search results so that the best performing content always rises to the top.

ART excels with popular and ambiguous queries where users enter only one or two terms. ART is robust to common typographical errors and learns implicit synonyms. When your Coveo index content includes secured items, ART queries the index to ensure to only recommend items the user performing the query is allowed to access.

In practice, ART boosts the ranking weight of recommended items so that they appear among the top search results.

CoveoART

The Coveo ML Query Suggestion feature recommends significantly more relevant queries to users as they type in the search box. The original usage analytics query suggestions are limited to top queries in which the typed characters exactly match a suggested query part. The Coveo ML Query Suggestions feature:

  • Identifies typed characters exact, partial, or fuzzy matches anywhere in any individual keyword appearing in any order.
  • Stems query suggestion keywords to remove duplicates.
    Offers the most relevant recommendations by ranking query suggestions considering:

    • The number of times the query was performed.
    • The degree of matching.
    • The query performance based on the Relevance Index and Click-Through usage analytics metrics.
  • Only considers queries performed at least 10 times and for which at least 5 had a search result clicked to eliminate outliers.

In the end, suggested queries are surprisingly tolerant to typos and get better as your usage analytics data set size increases.

coveoquerysuggestion

The Coveo ML Recommendation feature learns from your website user page and search navigation history to return the most likely relevant content for each user in his current session. The recommendations can be interpreted as “People who viewed this page also viewed the following pages”.

The recommendation algorithm is based on the co-occurrence of the events such as page views within a user session. When two events abnormally frequently co-occur within sessions, the algorithm learns that they are linked. When one event is seen, the model recommends the other.

coveomachinelearningrecommendations

Query Pipelines

Another great feature of Coveo ML is query pipelines. A query pipeline is an alternate set of rules or models that can be defined to modify queries.  You can take advantage of query pipelines when you have more than one search interface with distinct users and purpose and want to apply different rules or models for each.

The query pipeline rules can define:

  • Thesaurus entries – replacing or expanding queries with synonyms.
  • Featured results – items appearing at the top of search results when the query meets a specific condition.
  • Stop words – ignoring unimportant words in queries.
  • Ranking rules – modifying the order of results matching specified expression and condition.
  • Ranking weights – establishing the impact of ranking factors when establishing the order or results.
  • Triggers – establishing actions to be performed in the user search interface following an event when a condition is met.

The rules can be applied on the query before it is sent to the index (like thesaurus and stop words) or on the results returned by the index before they are sent back to the search interface.

The query pipeline models can define:

  • Automatic relevance tuning – optimizing search results relevance based on user search behavior.
  • Query suggestions – suggesting queries to users as they type in a search box.
    Recommendations – predicting and proposing the most relevant content for the current user in the current session.

Query pipelines are easily managed in the Coveo Cloud administration console for your cloud organization.

Coveo Cloud has impacted On-Premise

The Analytics module helps Content managers measure and improve what users are searching for and accessing. Essentially, events performed by end-users in Coveo search interfaces (and optionally in other web page types) are recorded by the usage analytics service in your Coveo Organization. You can then use the administration console Analytics pages to learn what end-users do by reviewing usage events, metrics, and dimensions in explorers and dashboards.

While the Usage Analytics module is available in on-prem the module has been deprecated and is no longer maintained by Coveo.  Instead, to use this module with an on-premise index Coveo recommend using Coveo Usage Analytics cloud service to connect to and monitor you on-premise search usage (see scenario 4 of the Coveo Cloud usage analytics use cases).

Coveo Cloud Concerns

Let’s address some concerns you might have when considering Coveo Cloud:

Q. How secure is data in the cloud?
A. Coveo is dedicated to the security, privacy, and integrity of their customers’ data. Coveo perform regular automated and manual application security tests and patch any potential vulnerabilities or bugs. Data processed by the Coveo Cloud is encrypted both at rest and in transit using best-in-class industry standards. Coveo’s commitment to security is affirmed by their SOC II examination and HIPAA compliance.

Q. Who owns the data in the cloud?
A. Currently the cloud solution is hosted in US East with backup in West so you will have to accept that your data is in the US. Although the product manager (Simon Langevin) confirmed that in the near future, hosting your local Elastic cluster and connecting Coveo Cloud on top of it could be a possibility.  Also Coveo is very transparent with their procedures pertaining to data retention and destruction that are in place and are also audited.

Q. How do you measure the success of Coveo Cloud ML features?
A. You can use the following 2 traditional marketing metrics to evaluate how successfully your community search connects users with the information they need to solve their specific issue: Click-Through Rate (CTR) – The percentage of users clicking on any link on the search results page. Higher values are better, meaning that users are more often opening search result items following their queries.

  • Click-Through Rate (CTR) – The percentage of users clicking on any link on the search results page. Higher values are better, meaning that users are more often opening search result items following their queries.
  • Average Click Rank (ACR) – Similar in concept to page rank, this metric measures the average position of opened items in a given set of search results. Lower values are better, as a value of 1 represents the first result in a list.

Coveo ML optimizes search results and query suggestions, and will, therefore, improve CTR and ACR metrics and contribute to increasing self-service.

Q. Can I separate my Production, Test and Development environment indexes in the Cloud?
A. Yes, all the Coveo for Sitecore cloud plans offer a Production and Sandbox cloud organization. For development, you can either purchase an additional organization or use trial organizations. The trial organizations are only valid for one month, but creating a new one is very easy.  This allows each developer to have their own organization to use during development.

Q. What happens if I go over the monthly query limit on Cloud Enterprise or Pro?
A. You should speak with your Coveo Client executive regarding the pricing structure for exceeding query limit as this depends in your license term. It is helpful if you have an idea of the anticipated monthly queries.  But don’t worry you won’t experience and degradation of service as Coveo will not block queries should you exceed the thresholds.

Resources

Coveo For Sitecore Edition Comparison

Coveo Machine Learning

Coveo Machine Learning Features

Coveo Machine Learning FAQ

Managing Query Pipelines

How does Coveo Secure your Data and Services

Best Practices for deploying Coveo for Sitecore Cloud

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s