No. It is a separate product and available to be purchased separately within Sprinklr’s Sprinklr Insights product family.

No. As both are separate products, MM&A does not consume mentions that are allocated to a customer for Listening.

Yes. If the customer has both listening and MM&A, the listening data is enriched with MM&A metrics, such as Impact, Influence, Web shares on Twitter/ Facebook/ Reddit, Similarweb Metrics such as Potential Reach and Global Rank, Publication, TV channel, Print Source Name, etc.

MM&A takes data from the global Sprinklr database which consists of data from the Twitter firehose (partial), News firehose, Facebook (public pages), Reddit, TV, and Print mentions that are pulled into the Sprinklr database.
All the Twitter mentions from verified profiles are considered for the story clustering algorithm whereas the entire Twitter firehose is considered for social shares estimation, i.e. if a tweet contains a news article URL.

On top of these, MMA does offer some additional sources –

1. Financial Times (Available by default)

2. 1500+ web sources from NLA (Available by default)

3. 5000+ print and web sources from Factiva (Available as a separate SKU)

4. TVeyes (Available as a separate SKU)

There are several features that are powered by AI in MM&A –

1. Clustering Algorithm: AI combines similar-looking news articles, social shares of these articles, TV, Print mentions, etc. in one story which reduces noise and helps PR teams to quantify the event better

2. AI-based two-line summaries for each story

3. News categorisations

4. AI-enriched sentiment for each mention

5. Smart search suggestions in the search bar on the landing page

6. AI enriched Brand and People Recognition

The title and content of a news article are the two main entities that decide the story to which an article will be added. For every article that enters Sprinklr, the title and content are analyzed. Closely matched articles are grouped together as a story. Also, mentions from TV and Print and tweets from verified profiles are considered for creating the story.

Any shares of the news article on social channels such Twitter, Facebook public pages, and Reddit, i.e. if a tweet/post contains a news article URL, are also grouped under the same story.

​There are three reasons why there could be two different stories with similar articles –

1. There is a probability of seeing two separate stories with similar news articles or social messages as the entire process is created automatically by AI. The clustering algorithms are tuned and improved on a regular basis.

2. Though the articles might look similar, the content might refer to slightly different news events. For example, “Elon Musk's Tesla launches tequila $250 worth tequila called 'Teslaquilla'” and “Elon Musk’s $250-Per-Bottle ‘Teslaquila’ Sold Out Hours After Debut” are similar but not the same news event.

3. The lifetime of a story is 4 days. Incoming messages are not added to the story if the corresponding story has been created before 4 days. We have this criterion to ensure that the stories are relevant and up to date. However, the social shares count (discussed later) are tracked for 7 days since the story was created.

However, if the client requires to analyze similar stories together, a story tag can be used to apply custom tags to stories.

​Out of all the articles grouped under a story, the article which has the highest popularity is selected. The image, headline, and summary are derived from it.

The title of the story would be equal to the headline of the selected article. Note that the title of the article is based on the title available on the online website or print media.

The summary of the story is derived as follows –

1. The entire content of the selected news article is broken into sentences and meaningful phrases.

2. Each sentence/phrase is given a rank based on its importance using a proprietary algorithm

3. Based on the significance scores, the top 2 sentences that are related (but not highly related) to the title are selected for summary

​The following is the methodology used in selecting the stories to be displayed –

1. Based on the keywords (story queries, story query tags, keyword lists, etc.) and the filters (such as domain lists, country, language, story tags, etc.) applied, the mentions in the MMA database are matched

2. All the stories that the mentions are part of, are displayed (on the home page or the story card widget)

3. For example, if 100 mentions are matched and these 100 mentions are part of 20 different stories, all 20 stories are displayed.

4. Please note that the title and summary of the story card are static and does not change with the filters applied (but it can be ensured that at least 1 message within the story matches the given criteria)

​Let’s take the below example of publication count to explain this:

1. The publication count metrics shown on the story card, i.e. the +912 publications is filter-agnostic (hence named as "publications globally")

2. The metrics plotted along with story cards are contextual to the filters, i.e. 461 publications

The difference in number is due to the fact that AI-clustering is able to match mentions better than keyword-based matches (topics/queries).
—> 1 will always be greater than or equal to 2
—> The more exhaustive the keywords and filters are, the less difference between these metrics will be.

​MM&A supports a wide range of sources namely News, TV, Print, Radio, Twitter, Facebook, and Reddit.

Over 400K news and websites are covered.

• Extensive Web News Coverage across 180+ countries

• Over 3M+ news mentions monitored every day

• Coverage across International, National, Regional news sources and Industry-specific media releases with sources being added at regular interval

• Enhanced monitoring of Licensed content from paywalled sites such as Politico, Reporter, The Street, etc.

• 100K+ news domains from the APAC region including major domains such as AntaraNews, Business World, Financial Review, Hankyung, NDTV, Nikkei, Pulse, Sputnik, The Hindu, The Mainichi, The Straits Times, Times of India, WA Today, etc.

• 150K+ news domains from the EMEA region comprising of important domains like ABC, Asharq Al-Awsat, Basler Zeitung, Dailymail, Die Welt, Diário de (Spain), El Mundo, Euractiv, la Repubblica, Mehr News, Metro, News24, SRF, Telebasel, Telecompaper, Times Live, etc.

• 160K+ news domains from the NALA region including major domains such as BusinessInsider, CNBC, CNN, ESPN, Forbes, FoxNews, USA Today, Washington Post, Wired, etc.

With regard to TV and Radio broadcast, the coverage is over 1,500 channels in all 210 US Nielsen markets, Canada, and the UK. These sources include:

• 1,000+ Local US TV stations including ABC, CBS, Fox, and NBC affiliates in all Nielsen DMAs.

• 80+ US cable channels including major cable news like Bloomberg, CNBC, CNN, Fox News, Fox Business, MSNBC, etc.

• 200+ US Radio stations.

• 200+ TV and radio stations in Canada including CBC, CTV, City, and global affiliate stations.

• 130+ TV and radio stations in the UK including BBC, ITV, and Sky content.

• 120+ US Spanish language stations.

On top of these, MMA does offer some additional sources:

1. Financial Times (Available by default)

2. 1500+ web sources from NLA (Available by default)

3. 5000+ print and web sources from Factiva (Available as a separate SKU)

4. TVeyes (Available as a separate SKU)

1. ​News: Text content as available on the online native site and whatever is legally permitted for crawling

2. Print: Text transcripts as printed on the offline print source (physical newspaper or magazine)

3. TV: Transcripts of the TV Broadcast via voice to text transcription or closed captioning technology

4. Radio: Transcripts of the audio recordings via voice to text transcription

​We do not claim to circumvent "paywalls", however, we do get data from certain paywalled sites. Moreover, we do have enhanced coverage of news which includes Licensed content from 5000+ major publishers, with 100+ paywalled sites already covered.

Paywalled/Premium content is handled by publications differently and so the data we get also differs widely. Below are a few possibilities –

• For certain paywalled sites, we get the full content. The reader can read the entire article on Sprinklr UI. This includes some premium paywalled sites covered via Dow Jones Factiva such as Wall Street Journal

• For certain paywalled sites, we get the full content. The reader can only read the first 250 characters article on Sprinklr UI due to the Fair Use Doctrine & copyrights. For example, NLA Web Sources and ft.com

• For certain paywalled sites (like thetimes.co.uk), the publication provides the first few paragraphs as free and hides the rest of the content behind a paywalled. In such cases, we get that snippet of data (whatever is visible on the native website as free). If the keywords are matched using that snippet, the article is shown in dashboards. From our UI, the user can click the permalink and be redirected to the website, where they can log in (if they have a subscription) and read the entire article.

• Certain sites are completely paywalled and do not provide access to crawl their data. In such cases, we would not be able to pull those mentions into our dashboards.

MM&A also offers premium/paywalled content from certain publications via Dow Jones/Factiva. Dow Jones/Factiva is available only for MMA clients as a separate SKU. For content related queries please refer to Dow Jones as a source for MMA.

​Our data providers provide data for our traditional sources like Online News with a general latency of less than an hour. More than 80% of the news data comes within the first hour of the article being published. Latencies for Print depends on the publication.

• ​The categorization of whether a mention is a news or a blog is taken care of when the data is extracted from the web

• Crawlers are written for each section of a site. So all articles that are extracted from a news domain, for example the domain "cnn.com" are labeled as NEWS

• In general, News data is set to be clean. For example, if there is a WordPress blog with an individual's personal thoughts, it would not be eligible for addition to the News content set.

• It is to be noted as News and Blogs are overlapping categories and there is some grey area. Some News sites have hosted blogs and some blogs have developed into full News sites. If a blog has Editorial oversight of content (a team of writers and editors), it might be eligible as News.

​Yes, as the database for both listening and MMA are the same, completing the source verification will add the source coverage in MM&A as well. Backfill of a newly added source is not possible in MM&A.

For further details please refer to Source verification FAQs

We constantly expand the number of web sources that are available for monitoring so that you do not miss a mention –

• We continuously add the web sources that are requested by our clients. Once a source is added, we ensure the data from the newly added sources are available for all our clients.

• An inhouse team that dedicatedly works on identifying new domains that are not already covered within Sprinklr. Once identified these sources are added for coverage.

• Sprinklr also works on building partnerships directly with global data vendors or publications to bring in exclusive paywalled/premium content. For example: Financial Times and NLA feed.

We obtain a certain list of print sources from our data vendors as part of our vendor partnership contract. Our data vendors in turn have contracts with print media houses to procure the transcripts of the print data. The availability of new print sources depends on API availability as well as on-boarding the new publishers depends on licensing and other legal aspects. Sprinklr can take requests for new print sources and pass on to our vendors. However, it cannot be added immediately or ad hoc as it depends on how the relationship is set between the vendor and publisher.

We also continuously work towards building partnerships with vendors specialising in print content to continuously expand print coverage.

​Currently, the AI algorithms required for creating stories in MM&A are trained using 30 languages namely English, Arabic, Chinese (Traditional), Chinese (Simplified), Croatian, Czech, Danish, Dutch, Finnish, French, German, Greek, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Norwegian, Polish, Portuguese, Russian, Serbian, Slovak, Slovenian, Spanish, Swedish, Thai, Turkish, and Ukrainian.

So, the MM&A database would have mentions only from these 30 languages.

​The language in which the title and summary are displayed in story cards can be controlled through the user language preference in "Users" settings.When language filter is selected as German on MM&A Home page and User Language set as German –

1. All story cards displayed would have German titles and summaries (there could be a few exceptions where we detect wrong languages).

2. All the story cards displayed will have at least one news article in German

When no language filter is selected on MMA Home page and User Language set as German –

• The story cards with at least one news article in German (the user’s language preference), will have German titles and summaries on them

• The story cards with no news article in German (the user's language preference), will have an English title and summary.

• The story cards with no news article in German (the user's language preference) or English (the default language), will have a title and summary belonging to one the other supported 14 languages.

​Different factors go behind in setting up the location for a news/blog site. Majorly, the location is tagged at the parent domain level. Assigning country to the parent domain based on the below points in priority:

1. Countries are determined based on the ccTLDs (country code top-level domains). For eg: If the publication is the dailymail.co.uk → Country=United Kingdom

2. Countries are determined based on the ccTLDs present in the URL (country code top-level domains). For eg: If the article URL is www.cnn.com/uk/this-is-the-article-name.html → Country=United Kingdom

3. Country can also be tagged based on the publication’s headquarters available on the "About" page. This is performed by human taggers.

4. There could be certain domains that are hard to code, in spite of the above heuristics. In that case, the best human judgement is taken to codify the country. Hence, classification errors could be present.

Also, it is important to know that ambiguous news sites are not assigned any country tag. This means filtering based on countries can result in lower mention counts.

​Sprinklr has a deduplication algorithm that dedupes based on URLs and UTM extensions and based on message content.

Users can leverage PR tags in order to remove any article from their analysis/dashboards. The PR tags can be used as filters at dashboard, section, widget or even at a story query level. For further details of using PR tags please refer to PR tagging in MM&A.

Similarly, story tags can also be used to remove mentions at a story level.

​Currently, the MM&A database has English language News mentions from April 2020 onwards. Other 30 languages (listed above) are available from Dec 2020.

"Story Query" can be created to pull stories that are related to the required keywords. It can be set up, similar to any other query available in Sprinklr, using operators such as OR, AND, NOT, etc. It is available under Settings in the left navigation of the MM&A App.

​Stories are broadly classified under different industry-based categories like Aerospace, Agriculture, Health, Hospitality, Technology, etc. Categories are assigned using our AI categorisation models based on the title and content of the article.

​The stories can be sorted based on several story-related metrics such as EMV, Impact, Influence, Publications, Overall Reach, Relevance, Shares, etc., and other variables such as Date Created.

​Story Tags can be used to tag similar looking stories. Then the Story tag can be added to a MM&A Dashboard to aggregate all the underlying data of the tagged stories.

Please refer to Metrics and Dimensions in MM&A to get further details.

Every story/mention is given a "Relevance ****score" corresponding to a particular search query or a story query. The relevance score is computed based on whether the keywords in the query are available in the title or summary or in the description of the mentions that match the applied filters.

The higher the relevance score, the higher is the probability that it matches the query. Therefore this is helpful to quickly sort and view the most relevant mentions/stories. The value in itself is not useful, but useful for comparisons. Currently the value can be as high as 10k+ (depending on matches) and there is a plan to index it.

"Influence" metric provides a measure of the importance of the publisher’s domain. Currently, it is computed using the MOZ Score. MOZ Score is a search engine ranking score that predicts how a data source like a news website will rank on SERP and also includes other metrics. Higher the domain’s MOZ score, higher is the influence score.  The score is out of 100.

The Earned Media Value or the EMV metric of a PR content measures the dollar value that the PR content could generate when it gets featured by a publisher or outlet. It gives a measure of the monetary worth of the exposure/promotion received by the brand. The formulae for EMV is proprietary and is dependent on four factors namely –

• Media Reach: It’s the circulation numbers for print, viewership for Broadcast (from Nielsen), and Unique Visitors per Month (UVPM) for Online News

• Word Count: The number of words, or the length of the article

• Ad Rate: Cost to place an ad in a certain publication based on the amount of space the ad will take up. This is provided by our data partner who maintains a database of this information (average ad rate based on the position of ad in website/print). Sometimes it is based on estimates. The value is updated quarterly.

• Source Rank: Editorial ranks that are applied to news sources. It is a source-level categorisation with 5 levels

  • International, National, and business news sources

  • Regional sources

  • Industry-specific or top organization media releases

  • Local sources

  • Non-news sources such as message boards

• If we don’t compute EMV for a particular domain, the value is shown as zero.

  Another popular way to estimate EMV of web sources if by using a Custom Metric. Typically PR teams have a formula to calculate EMV (typically based out of Reach numbers) and we could replicate that using custom metrics feature. One of the popular formula used for EMV metric is EMV = Similarweb's Total News Media Potential Reach * 2.5% * $x.xx, where

  2.5% is the factor assumed to better estimate the number of audience who actually viewed the particular article published in the domain

  $x.xx is the dollar value that the brand places for every reader who is reading that article.

We are unable to get Media Reach values for certain print or TV channels. We try to continuously improve. Please reach out to the Product team if you find an important source with Media Reach=0

Total News Media Potential Reach metrics is calculated at domain level and hence the metric would be the same for all the articles from a news/blogs domain.

We partner with Similarweb, a leading website-performance measurement partner for this metric. It is possible that for some publications Similarweb does not provide this metric, and for such publications the value of this metric will be 0.

Also please note that the metric enrichments are available from Feb 2022 onwards. (i.e. the metric would be assigned a value of 0 before Feb 2022)

​The "Web shares on Twitter" metric counts the number of times a news article is shared (its URL) on Twitter. We also track the number of times the URLs are shared on Facebook and Reddit as well. For Facebook, we monitor over 100K+ public pages, where the news URLs can be shared.

A separate section for analyzing the social shares is available in the platform which provides the trend of social shares, sentiment analysis, channel distribution, etc.

Yes. Check the social section after clicking on a story (Story Dashboard). We can see the top advocates, detractors and also understand which article is getting the highest shares. Similar widgets can be built in custom dashboards.

Yes, domain lists can be built, where users can group publications of interest. For example, Domains that syndicate news can be grouped separately and analysed. Or if there is a use case to track the performance of tier-1 vs tier 2 publications, domain lists can be used.

​The dimension "Journalist" in MMA is pulled automatically using the author details available on the web domains. It is not linked with the respective social media profiles of the journalists. If social profiles of interested journalists are available, listening and analyzing those profiles can be achieved using domain-based listening.

​The various components used to derive an Impact score for a news article/social mention/TV coverage are:

• Influence

• Shares

The score is out of 100 and is based on the weighted average of scores obtained from the above two factors. More factors are planned to be added in 17.10.

​Yes. Story Analytics has to be used as the Data Source while creating MMA widgets in Listening Dashboards. Make sure you apply the required Story Query as a Widget-level filter, as the Story Analytics Engine runs on Story Query.

You could also plot MMA dimension/metrics/stories using Listening as Data Source. This is because, as explained earlier, if the customer has both listening and MMA, the listening data is enriched with MMA metadata as well.

​We have a partnership with a vendor who specialises in collecting and maintaining the contact information of the Journalists and Publications. The contact information of the profiles is updated regularly by the vendor. We sync our database with that of the vendor’s in real-time to keep the profile information as updated as possible.

​There could be some journalists/publications for which the vendor does not have the contact information and hence it is not available within the Journalist & publication discovery module. This could be due to multiple reasons, such as

• Journalist has not given permissions to render their contact information

• Business-related contact information is not available for the profile

Currently, users will not be able to override the existing information available in the profile dashboards, but in order to record some critical updates, collaboration widget in the overview tab can be used.

Currently, it is not possible to contact journalists/publications within Sprinklr. Users can gather information from Sprinklr and privately contact the journalists/publications using their existing communication tools.

Adding journalists/publications is a continuous process as undertaken by the vendor. As we regularly sync our system with the vendor’s database, we have all the updated information w.r.t all available profiles.

The vendor captures and stores profile information in a compliant manner and according to GDPR guidelines. The vendor does not store any sensitive personal data about the profile unless the data is put into the public domain (such as on public Twitter timeline) by the profile itself. Most of the personal data is obtained from the profile directly or from publicly available sources such as:

1. Social media profiles

2. The articles the profile have published

3. The videos profile has shared

4. Profile’s own website

5. Profile’s employer's website

6. Other publicly available online sources

The overall coverage depends upon how each tool is gathering the journalist/publication data. There could be a difference in database if they are using a separate vendor altogether or are managing the data in house. However, following are the key advantages of using Sprinklr solution –

1. Detailed Twitter content analysis of journalists to help filter journalist who write about certain themes of conversation, have significant earned engagement, drive positive sentiment, etc.

2. Detailed follower analysis to understand if the journalist will suit the target audience of a particular campaign

Yes, Sprinklr allows you to send newsletters to multiple recipients directly via Sprinklr platform. With Sprinklr’s newsletter you can communicate the top event using Listening and Story Analytics data source. Please note that only Listening and MMA clients have access to newsletters. For further Information, please refer to Introducing Newsletters to Sprinklr Insights.

Yes. With Sprinklr, you can now customise your newsletters to align with your communication strategy. Sprinklr’s newsletters allow you to –

1. Customise the branding of your newsletters to align with the branding guidelines of your organisation

2. Customise the content of your newsletter messages to drive key insights using manual newsletter.

3. Add additional insights and internal announcements to complement the newsletter messages using Text and Title widget.

No. There is no restriction with respect to the number of recipients that you can add in your Sprinklr newsletter.

With Sprinklr, it is now possible to track the performance of your newsletters and understand the various reception metrics for your newsletters via Newsletter Analytics.

Various email analytics are available for tracking including –

1. Number of Recipients

2. Number of Emails sent

3. Unique Opens

4. Open Rate

5. Unique Clicks

6. Click Rate

7. Unsubscriptions

8. Unsubscription Rate

Newsletter analytics will be available for the newsletters created and distributed post 23rd August 2022. Also, please note that the Newsletter analytics are currently available at a distribution level only.

No, the recipients won't be able to read full paywalled articles within the newsletters. Due to fair use doctrine policy and copyright rules, only the first 250 characters are shown in the newsletter messages.

Once the user clicks on a newsletter message, they are redirected to the native site of the article. If the user has access to any paywalled publication, they can use their personal credentials to login and see the paywalled content.

However, Sprinklr is planning to introduce a new feature to allow Dow Jones content within the newsletters. (planned for 18.1 release)