B2B technology is opaque and complex. B2B sales processes rely on personal relationships rather than relevant facts. B2B analysts share their opinions.
B2B technology research is too important to be opaque and complex. It guides hundreds of billions of dollars in annual spending.
Editorial policy
Our mission is to help businesses make better technology decisions thanks to transparently-shared, relevant insights. Here is how we strive to achieve that:
Research process
AIMultiple team follow the 6-eyes principle: Every AIMultiple is read by at least 3 industry analysts.
- Industry analysts research topics and prepare their findings in a draft. They are credited as "Researcher" in the article bylines.
- The draft is reviewed by another industry analyst. The reviewer is not credited in the bylines. Their focus is to ensure that the article relies either on the latest and most reliable sources or our team's own experience with the technology.
- Principal analyst, Cem Dilmegani, reads and finalizes the article. He is credited as "Author" in the article.
Select articles are also read by AIMultiple Business Leader Network members and their input is used for fact-checking and get more user input. Business leaders
- Are users or managers of the users of the technology solution they review.
- Can not be current or former employees or partners of technology providers in the areas where they review AIMultiple articles to minimize conflicts of interests
Guiding principles for AIMultiple research
Simple and direct
B2B websites and analysts frequently use indirect language and jargon.
We aim to be as concise as possible, directly explaining key points.
A theme may include many topics that may be relevant to different audiences. These will be explained in different sections and users should always navigate to the relevant section for them with ease.
Relevant
Sharing too much is worse than sharing too little. Business users have competing priorities and AIMultiple research needs to aid their decision making by sharing only relevant information.
Data-driven
AIMultiple doesn't endorse any products or services, it shares relevant information about them that is based on verifiable facts.
Ethical
AIMultiple team will remain independent and aim to avoid conflicts of interest to remain objective. We strive to achieve the highest levels of transparency.
Conflicts of interest that can't be avoided (e.g. AIMultiple commenting on our customers’ services or products) are highlighted clearly in our articles. For more, please see our ethical commitments.
Criteria for relevance
AIMultiple articles only focus on points that matter.
Business technology products are complex, having hundreds of features that can be evaluated in numerous different ways. Products are also constantly evolving, making evaluations challenging. AIMultiple will not focus its assessments on
- Features that are provided by some solutions in a domain. Table stakes in a category don't drive businesses to choose one product over another
- Statistically significant differences. Most benchmarks are completed with a limited data set and in most cases, minor differences are not statistically significant. AIMultiple does not publish such differences that are unlikely to be reproduced and focuses on statistically significant differences.
AIMultiple is constantly improving its research methodology and commitments. Please reach out in case you have comments or suggestions.
Data sources
Vendor influence reaches every part of the B2B information ecosystem including the media, industry analysts, influencers and review platforms. To ensure the validity of our data, our data sources need to be credible satisfying these criteria:
- Clear: The reader should see where how the underlying data was accessed
- Verifiable: A user should be able to verify the data behind our insights.
- Relevant: There is too much data but few are relevant for a specific analysis. We will focus on the relevant ones.
If such data is not available, we will share our findings which will be as a result of one of these channels:
- Our team's use of products or services
- Our benchmarks
- User or expert interviews
In the absence of verifiable data, AIMultiple team may also share information that is not verifiable. In such cases, AIMultiple will clarify in the article that this is a claim not a verifiable fact.
These are the typical types of data used in AIMultiple research and their update frequencies. If AIMultiple team uses different data sources for these fields, they will be clearly identified within the article:
Market presence metrics
We typically rely on these metrics since they are correlated with higher revenues for B2B technology firms and can be publicly verified.
Number of employees:
- Rationale: For companies competing in the same segment, there is a correlation between a company’s number of employees and its revenue. This correlation is not a perfect one and we are aware of its limitations:
- Companies with multiple products tend to have more employees than companies that focus on a single product. This may not mean that their products are necessarily better than more focused companies' products.
- More efficient companies may be able to deliver better products with the same number of employees than less efficient companies.
- Workforce seniority is another important factor not reflected in number of employees.
- Number of employees is one of many cost metrics. A small workforce with access to great resources (e.g. thousands of GPUs) may deliver a better product than a large workforce with access to fewer resources.
- Source: Linkedin
- Update frequency: Monthly
Number of reviews and average review:
- Rationale: For products competing in the same segment, there is a correlation between a product's number of reviews and its revenue. This correlation is not a perfect one and we are aware of its limitations:
- Some products run review campaigns with review platforms and others don't which leads to different numbers of reviews for products with similar popularity.
- Free versions can boost reviews but may not improve customer experience for paid users.
- Reviews can be incentivized, auto-generated or written by the vendors themselves or their partners.
- Source: Top 2 pure-play review platforms. For now, we are not including review platforms that are owned by industry analysts. They will be included as we expand our data collection efforts.
- Update frequency: Quarterly.
Features
- Source: Our benchmarks, vendor websites, reviews or partners of vendors
- Update frequency for
- Vendor websites: Daily. However, not all of the information on AIMultiple is covered by this daily update model and our team is working hard to ensure that all critical information on AIMultiple is daily updated. If daily update is not available, the updates will be at a minimum annually during updates.
- Reviews or partners of vendors: At a minimum annually during updates
We may need to count features to give an impression about feature completeness about products. In such cases, we'll follow this method:
- Binary (i.e. true or false) features are counted as 1 if they are true. For example, for the feature: "ISO 27001", if the product has this certification, this feature would be counted as 1.
- Numerical features are counted as 1 if we have identified the numerical value
- Features that can take on multiple values are counted by the number of values that the product supports. So for example, if the product provides the values "SAP" and "Dynamics" for the feature: "ERP Integrations", this is counted as 2.
Pricing
Prices may be displayed in dollars without including cents. Therefore, a price of $14.99/user/month would typically be displayed as $15/user/month.
Source and update frequency are the same as outlined above in features section.
Open source
To be included in articles that highlight solutions in a category; open source solutions need to have been updated within the last 6 months. This is checked during updates and outdated open source solutions are removed from our pages.
Corrections and updates
Technology products and services are ever changing. To be worthy of our readers' trust, AIMultiple team has designed a regular update system:
- All AIMultiple research articles are reviewed at least on an annual basis. Most of our research is updated on a quarterly or monthly basis.
- Constantly changing information such as product prices and feature availability are updated either daily or somewhere between daily and annually. For more details, please see the data sources section.
Third-Party Content
All research articles on AIMultiple are prepared by our own team. AIMultiple doesn't host any third-party content with the exception of:
- Some of the videos shared in research articles
- Some of the images shared in research articles. Source of images will be highlighted in the article
Version
This is version 1.2 of AIMultiple's research guidelines published on July 3, 2024. It will be implemented in all AIMultiple pages by January 6, 2025.
