Here are the statistics from scraping 1-100 search engine results pages that rank for "drug rehab" in the USA. (scroll to the bottom for the TL;DR)
Birds of a feather flock together
Before we start things off, yes, correlation does not imply causation, and yes, this survey is too small of a sample size. But what it does show us is similarities amongst the top-ranking pages that rank for a specific keyword.
You can use this data as a model on where to set your goals, breaking it down into smaller pieces, giving yourself a better opportunity at ranking first in your industry.
Ranked Page Types
I see a healthy mix of homepage and article pages ranking for "Drug Rehab." Homepages naturally have more backlinks, which gives it an advantage. Still, this does not stop you from ranking article pages.
Dot COM and dot ORGs dominate the list; this reveals people click on recognizable TLDs (top-level domains) that they trust and are used to seeing. Hyphened URLs and newer extensions won't hurt you in the SERPs from a ranking factor, but it can affect you at a user experience level (e.g. drug-rehab.xyz looks spammy compared to drugrehab.com).
Since this was a US-based scrape, it pulled US search results. The results would differ for international SERPs, e.g., dot CA's in a Canadian based query might dominate the SERPs for a specific keyword.
SSL Certificates are vital to have; the 2% of non-SSL sites that ranked appeared to have been around so long that they continue to found in the SERPs, or the SSL expired, and the index has not updated (it can take weeks for months for the index to fully propagate).
It doesn't appear which DNS provider you use affects rankings. GoDaddy most likely made the list because it's such a popular registrar, not for performance and availability.
98% of websites that ranked only have one IP address per website —dedicated IPs signals premium hosting (Virtual Private Server [VPS], Dedicated Hosting). Shared (cheap hosting) would have multiple websites on one IP address.
Investigating the two websites (2%) with two websites sharing the same IP; one was a sister medical website, the other had been removed, but was still showing up in the scrape cache.
Insisting dedicated IPs does affect rankings contradicts John Mueller's claim that cheap hosting does not affect rankings. I would like to meet in the middle and say it might not be IP range, but dedicated hosting matters for performance. So, technically, you could rank with shared hosting, but the technical hurdles you would need to overcome would be enormous.
Content Management Systems + HTTP Server Head
WordPress dominates the CMS industry, so no conclusive data can be pulled from this, except that webmasters prefer WordPress.
HTTP Server headers might signal that faster is better, the fastest technologies appear to be dominating the space, but this could also signal what might be the most widely used.
Word Count is important, but a ten thousand word article isn't necessarily the secret to success. A twelve hundred to two thousand word article will be sufficient to help you rank.
"Word 1" represents the primary keyword found on pages. "Treatment" at 29% indicates informational search intent, answering questions on how people can get help for addiction. Keyword stuffing "Drug Rehab" on a page would hurt your rankings as its sparsely used.
Words without intent is spam, so remember to fulfill users' search intent; that is why people use Google - to have questions answered, so remember to add value to the end-user, and that will guarantee success.
Keeping a positive outlook in articles appears to be helping, maybe not with Search Engines, but with the end-user.
Flesch Kincaid Reading Score
The Flesch Kincaid scores 40s is difficult; 70s is easy. Scientific papers score in the 40s, so it's good to keep your articles medical + scientific. The average Flesch score for Harry Potter was 72.
Everyone has an About + Contact Page, Social Media Links, and Contact emails attached to their website. I don't see it as a ranking factor, just standard practice.
Search engine optimization metrics
SEO metrics developed by Ahrefs, Majestic, and MOZ are company driven metrics, not a metric Google shares. They hint at what Google might be looking at, but it's to be taken with a grain of salt.
That said, you will want a score of at least 20-30 to get yourself in the runnings to rank, but it's not a guarantee.
Backlinks - After analyzing the backlink scrape, I noticed some sites with less than ten backlinks outranking ones with millions (literally). So I dove a bit deeper in Ahrefs and noticed a trend of high authority backlinks (higher UR - URL Ranking) outranking numerous low-quality spammy links (UR 0-10) even if that same backlink has a high DR (Domain Rank). These high-quality links had contextual anchor text and from reputable sites in the same niche.
High Domain Rank should be paired with a high URL Ranking to be useful; a backlink from the New York Times (DR 94) would need a UR of +20 to be genuinely effective.
TL;DR - Key Takeaways
- Choose a recognizable/trustworthy domain name
- Enable SSL
- Have only one IP address per website
- Pick a fast Server/Hosting (if not fast, heavily optimize for speed)
- Article length of 1200-2000 words is average & acceptable to rank
- Focus on informational search intent by answering people's questions + being helpful
- Write proper medical copy (Flesch Kincaid Reading Score: 40)
- Aim for a DR/UR of 30+ score in Ahrefs
- Obtain relevant, contextual backlinks (quality is better than quantity)