I started getting questions about HostingReviews.io shortly after launching. Instead of answering the same questions over and over, I decided to put answers in one place (hence this FAQ). Transparency is a big deal with this project so this is where I'll lay it all out. If you haven't yet, also read the About page for more information on how things work. Please Contact me if you have any questions.
Is it a problem that people are more likely to tweet something negative?
While it is true that people are more likely to tweet a negative experience, this does not affect rankings. The reason is that scores are calculated by the ratio of good to bad tweets. The only effect that an increased likelihood of negative tweeting has is that scores are lowered across the board, for all hosts. This is why conclusions are based on rank. Hosts are classed as "Average", "Above Average" and "Below Average". For example, a host with 40% positive tweets may be ranked #25 of 50, so they are regarded as "Average" compared to others even though their actual score is on the lower end. Charts do not simply show a score. They show the host's score compared to the average score of all hosts. Comparison is the key with this project.
Are all tweets included?
Twitter's steaming API does not push all tweets into the database. Additionally, the queue of collected tweets that need to be processed sometimes becomes so large that it is necessary to randomly eliminate a portion in consideration of the manpower that is available. HostingReviews.io is like a researcher taking a sample large enough to draw a reasonable conclusion from. It is neither practical nor necessary to include every single tweet.
This means you cannot tweet with the expectation of inclusion on this site. This is not a submission site but rather a research project based on a sufficient sample.
Can I submit a tweet that was missed?
This is a good thought but can skew scores and disrupt rankings. Imagine if some people submitted only positive or only negative tweets for certain hosts. Only including tweets that Twitter's streaming API pushes into the database ensures equal treatment of all hosts. As a safeguard, no submissions will be accepted. Read the answer to "Are all tweets included?" above for more on this.
If you are a host and are concerned that not all tweets about you are included, know that both good and bad tweets are excluded (and included) without discrimination. All hosts are treated equally in this way. Enough tweets will eventually be collected for a host to rank (it currently takes a minimum of 10).
Is sentiment analysis automated?
No. Automation is used for collecting tweets that mention the hosting providers in the database. Each tweet is then read and processed by a human (either included if relevant or otherwise discarded). This provides better accuracy than automatic sentiment analysis which is a very difficult thing to accomplish.
The rankings don't match my experience. Are the scores wrong?
What you see on this site is an overall picture of what the masses are saying on social media. Rankings and scores will not match every person's own experience. That’s where the advantage is. You can ask one person about their experience with hosting and roll the dice to see if you end up just as pleased as them or you can take a look at what many people think and increase your chances of finding a host that works out.
How can you do this for free? It must cost money to run.
My purpose is to help people and make money. This site currently does not generate revenue. My initial focus is to build something useful and make people aware of it. I will experiment with monetization after it is established. One thought I have is to add a sidebar advertising products that people who buy web hosting are often interested in (web design training, WordPress themes, mailing list software, etc.). Another idea that I haven't given much thought to yet is to show Google AdSense units.
My theory is that a hosting review site with no affiliate links will be more trusted and therefore have a larger audience. A review site with affiliate links will probably earn more per visitor than one using an alternative method of revenue generation. But, if having no affiliate links means more traffic, it may be possible to earn just as much with an alternative method. To do so without risking trust and independence is an exciting prospect.
Are you opposed to affiliate marketing?
I hope people don't misinterpret what I've said about many web hosting review sites being primarily focused on commission as meaning I am opposed to affiliate programs and affiliate marketing in general. This is not the case. I run a very restrictive affiliate program with my own business and I participate as an affiliate in various programs for other projects. It is individual companies and people who set standards for their marketing practices. Some do better than others so it's not fair to make a blanket statement saying all are bad. It's only the ones that operate with low standards that I am opposed to.
Since so many web hosting review sites are questionable, I don't want to cause people to doubt the usefulness of this site by using affiliate links. See my thoughts above on alternate methods of monetization that should have less of an effect on trust and independence than affiliate links.