Back in August 2014, Google confirmed that is would be considering HTTPS protocols as a positive signal in its algorithm. The rationale, Google claimed, was that HTTPS was considered to be a sign of trustworthiness and security, which would protect users.
Google said from the outset that, as a ranking factor, HTTPS would carry significantly less weight than other known ranking factors, but it is a factor nonetheless.
We’re now just other three months since that announcement, so what impact has HTTPS had on search results, and can we identify any trends of correlations that demonstrate just how much importance Google is placing on a secure connection?
We have interrogated our unique dataset, Roadmap, which consistently monitors search engine rankings and ranking factors, to assess the real impact of HTTPS.
Brands are switching to HTTPS
At a very top level, we can see that there is an increasing trend for HTTPS results to appeal in search results, even if the increase is very slight. Just 5.52% of top 100 ranking pages in May 2014 featured a HTTPS protocol, but this figure has increased to close to 7% (6.93%) in November 2014.
Of course, it’s very difficult to draw any sort of meaningful conclusion from such a small increase, but remember that this is a ‘whole of market’ comparison, covering all 100 positions within all of the key markets that we analyse.
What is happening on a sector by sector basis?
It is when we look at individual sectors that we start to see an increased number of results featuring a HTTPS protocol. In particular, the trend stands out in highly regulated sectors.
The financial sector is one example of an industry where security and privacy are key consumer concerns. This appears to be reflected in the proportion of HTTPS pages ranking in the search results, with the proportion of ranking pages growing from 9.31% in May 2014 to 12.46% in November. In the gaming sector, we see a similar story, with the proportion of ranking pages with HTTPS protocols growing from 10.58% to 15.14% in the same period.
Getting on the first page
The above at least points to some form of trend, although we’re dealing with small increases that don’t, in isolation, indicate that Google is placing a huge emphasis on HTTPS as a factor on whether a site deserves to rank. But when we look purely at the most prominent positions, we start to see more of an indication that HTTPS could be playing a bigger role.
As an example, we have taken a look purely at the sites raking in the top five positions only in the financial services sector. In this case, we see an increase from 10.51% of ranking pages in the top five in May, to 17.41% in November – close to one fifth of top five search results in the financial sector.
That data looks at the financial sector as a whole, but what about specific markets within the sector? Again, we see some interesting trends.
Google has gone on record in saying that the short term (or payday) loans market is one that it has been monitoring and looking to clean up for some time, so is Google using HTTPS as a mark of authority and trustworthiness here? It would appear so.
The first graph represents the proportion of HTTPS results for the top 100 positions in the short-term loans vertical. Here, we see that more than one quarter (25.23%) of ranking pages are HTTPS. That is a notable proportion in its own right.
What is particularly striking however is the state of the SERPs when we look purely at the top five positions within the same vertical.
More than 70% of pages in the top five positions in November are behind a HTTPS protocol, compared with 36% in June. That’s a very notable increase.
Looking at the growth of HTTPS domains by position in this vertical in particular, we can begin to see an argument that secure protocols are playing a role in the market.
The trend lines infer that secure protocols are earning greater prominence in Google search results. Over the last six months, less than half of the pages in the short-term loans market ranking in positions ten and below have been HTTPS results, but as we start to move onto the first page, the picture is very different, particularly since November. HTTPS pages accounted for 83% of position one results in this vertical, with 60% of position two results and 76% position three.
But what about the lesser-scrutinised markets?
The payday loans market is a very easy target for this sort of analysis. It is an industry that Google has had concerns about for some time and on the face of it, secure connections represents an easy and objective metric when determining whether a site can be trusted.
But we do see similar trends in the less scrutinised markets that we track (finance, gambling, insurance, legal, retail, and travel), albeit not necessarily on the same scale.
We have looked at the legal sector, which for the purposes of our analysis includes markets for business services, clinical negligence, personal injury and wills and probate, and we can see a very significant weighting towards secure protocols for results in the top three positions.
It appears that the ‘wills and probate’ section of the market is what is really driving this trend, with the proportion of HTTPS domains in the top three search results significantly out-stripping other verticals within this sector.
In fact, it would appear from our data that it is only the travel sector where HTTPS is not a strong factor behind whether a page ranks in the top three positions, with the proportion of secure protocol results ranking in the top three positions actually falling from 2.52% in May to 1.33% in November.
We could explain this with a number of possible factors. The travel sector is one that is heavily influenced by aggregators, more so than other sectors in this dataset, and SERPs tend to be more focused on the delivery of content, rather than transactional sites. A lack of competition is also likely to be a factor, with the SERPs tending to be dominated by a very small number of established brands. The similarly low proportion of retail pages with HTTPS protocols could also suggest that site size is a potential factor.
Is depth of range a factor?
Retail and travel websites often have a much broader depth of pages and content than other transactional sites. Every hotel, excursion, product and catalogue item has its own page, giving Google much more content to work with.
We can see parallels with this specifically in the gambling sector. Sports bookmakers, for example, tend to have the deepest websites within the industry, largely because every individual sporting is listed on its own indexed URL. Because of the nature of sporting events, the lifetime of those URLs is also very short, guaranteeing fresh content. Poker websites can also vary in size, with a very strong emphasis on content (such as game advice, tournament information, etc)
On the other hand, bingo and casino websites tend to be significantly smaller. The games served on these sites tend to remain on a single static URL for a considerable period of time and as a result, Google has fewer pages to index.
This is perhaps reflected in the proportion of HTTPS pages ranking in SERPs, with the proportion significantly greater in the top three search positions for both bingo and casino.
This suggests that Google isn’t necessarily looking for HTTPS protocols any more than it is any other ranking factor. Instead, Google is merely looking for the most relevant content for the query – no real surprises there. To suggest that HTTPS is a way to increase your chances of ranking is extremely simplistic, but in sectors and industries where smaller sites and fewer ranking pages are the norm, it is clearly an indication of trustworthiness.
What can we learn from this?
This analysis was always going to be one that looked for indicative trends, and I think that we have identified a couple of examples where there appears to be an indication that Google may favour secure connections. This analysis does tell us that a greater number of HTTPS sites are now ranking higher than ever before, and that this is a larger consideration in some markets than others. Causation is obviously a concern with any form of analysis of this type and it is a highly subjective area. Whether those sites were already ranking in high positions is a different question but, even if high ranking sites have simply switched to HTTPS, doing so definitely hasn’t had an adverse effect on their rankings.
We can see some indication that changing to HTTPS has a positive impact (although we certainly wouldn’t expect it to have a huge impact on ranking movement as an isolated factor). When looking at the two most recent months, the data very slightly trends towards the top right quadrant, indicating that more domains have benefited from moving to HTTPS than not. Either way, we certainly wouldn’t expect it to have a huge impact on ranking movement as an isolated factor.
Could all those sites experiencing positive movement also have made other changes which positively impacted their ranking? Absolutely! There is no way to definitively isolate a signal’s impact. Looking at our larger dataset, the influence across any isolated ranking signal is typically minimal; it is the cumulative impact that starts to show change. Without sharing the data across hundreds of signals, what we can say is that whilst the influence of HTTPS may be minimal, it is currently positive.
Protecting users is very much in Google’s interests so we would expect to see these trends continue in the coming months and beyond. This will undoubtedly filter out some of the more unscrupulous operators from the marketplace who will not or cannot meet the criteria and the expense of certification for secure protocols – particularly in those highly scrutinised markets.
This article was originally featured by Econsultancy.