Unless you’ve been living under a rock for the past week, you’ve probably read or heard something about #TheDress. The debate went viral last Thursday, resulting in a massive influx of publicity for the retailer Roman Originals and an overnight 347% increase in sales for the garment in question.

Yet, beyond an initial spike in product interest, what will the legacy of #TheDress be for the brand? Additionally, based on previous academic research, what can we learn from the stories success?

Using these questions as a starting point, this post will attempt to explain:

  • Why this story went viral
  • The potential long term SEO benefit to Roman Originals

What Makes a Story Go Viral?

Like most SEOs, i’ve faced the disappointment of a well planned out piece of content falling flat — and have had to attempt to explain to a client why it didn’t perform as anticipated.

The harsh reality is that the sheer quantity of quality content being produced has made it extremely difficult to make a significant impact, and often the difference between successful and unsuccessful campaigns can feel like pure luck.

Yet, whilst luck certainly plays a part, certain types of content do seem to perform consistently better than others. This means that it should be theoretically possible to anticipate the types of content that will be successful, by comparing them to a specific criterion.

In his book, ‘Contagious’, Dr Jonah Berger lists six factors which can serve as a framework for successful viral content. These are:

Social Currency: People like to share things that have some sort of benefit them. For example, reinforcing their views or values, or helping them to be perceived in a certain way.

Triggers: Effective content often incorporates stimuli; external triggers that remind people of the original piece and help to keep it on peoples minds.

Emotion: If something evokes a strong emotive reaction, we are more inclined to share it — although some emotions can be less beneficial.

Public: We have a tendency to imitate the behaviour of others. For example, content that is observable and inclusive

Practical Value: We all crave information that is valuable and will make a difference in our lives, or the lives of those we care about. Hence the popularity of practical content with clear utility.

Stories: As human beings, we have always disseminated information through storytelling. Wherever possible, content should attempt to tap into this love for stories by creating a compelling narrative.

In my mind, #TheDress debate tapped into at least three of these factors:


The dress appeared to be completely different colours to different people, eliciting strong emotions from those on either side of the debate. These included: awe, surprise, confusion and frustration.


The division in opinions created by the garment, caused a very public and large scale debate that everyone could freely participate in.

Practical Value

The story was both fairly unique and highly inclusive, presenting clear value in sharing the story with others.

So as SEOs what can we learn from this? Primarily that content success isn’t always determined by budget. Sometimes — like in this instance — you can get lucky, regardless of the effort or planning you put in. Nevertheless, if you use the above framework as a litmus test for your ideas, you stand a much better chance of achieving success.

SEO Benefits

Now we’ve covered the factors that helped the story go viral, it’s time to analyse the SEO benefits that Roman Originals may see from #TheDress, due to the large quantity of high quality links the story has accrued.

On 27/2/15, the website only had 508 linking root domains; a fairly small number considering the competitiveness of the fashion vertical. As of 4/3/15, this had already increased to 1213 domains, a rise of 138.8%.

The vast majority of these links were also of a very high quality, causing the websites Majestic TF to increase from 17 on 27/2/15, to 27 on 4/3/15. Some of the best links included:

URL Domain TF Domain CF
http://independent.co.uk/life-style/fashion/news/the-dress-actual-colour-brand-and-price-details-revealed-10074686.html 78 76
https://yahoo.com/health/is-this-dress-blue-and-black-or-white-and-gold-112194158507.html 92 97
http://metro.co.uk/2015/02/27/this-is-why-your-facebook-and-twitter-feeds-are-full-of-chat-about-that-dress-this-morning-5081515/ 36 51
http://myfoxdc.com/story/28219686/real-color-of-the-dress 45 46
http://cbsnews.com/news/blue-and-black-dress-is-not-white-and-gold-uk-retailer-says/ 73 80
http://telegraph.co.uk/technology/social-media/11439084/So-what-colour-is-this-dress-The-science-and-theories-behind-dressgate.html 79 81
http://mashable.com/2015/02/26/buy-blue-black-white-gold-dress/ 77 84

Penalty Prevention

As well as the direct ranking benefits the links will likely bring, another perk is reduction in probability of the Penguin algorithm negatively impacting on the website. Or alternatively — if the domain is already affected — a higher chance of escaping the algo, without submitting a disavow file.

Whether or not it is possible to escape Penguin by building or attracting high quality links is a topic that has been discussed in some detail before. The general consensus is yes, because Penguin works by looking at ratios of good and bad links, before penalising websites on a gradient of impact. As such, it is theoretically possible for an affected domain to recover if these ratios change.

The influx in high quality links from news sites covering the story has resulted in a massive increase in Trust Flow for the domain, which is a good indication that this ratio has shifted somewhat. More interesting, however, is the change in the proportion of commercial vs branded and generic anchor texts.

Roman Originals anchor text profile previously showed relatively high levels of links using commercial anchors, including target keywords like: ‘evening wear’, ‘evening blouses’, ‘evening wear dresses’ and ‘womens clothing’.

The recent influx of links has massively improved this ratio, as most of the websites linked with branded or generic anchor texts. Consequently, by 4/3/15, the profile looked like this:

As anchor text ratios are one of the primary factors Penguin looks at, it stands to reason that this increase in branded and generic terms may both serve to prevent any future penalties occurring, or help the brand to escape any current penalisation. In this instance, the former scenario seems more likely, as Semrush shows no large-scale drops in organic visibility.

Note: Without seeing webmaster tools, there is no way of knowing whether or not these links have already been disavowed, in which case this would all be a non-issue. Nevertheless, the above is still an interesting example of how a viral piece of content could potentially lead to a website escaping an algorithmic penalty.


Although it was largely accidental, the success of #TheDress is still a good case study for the ways in which viral content can benefit a business. The challenge for SEOs and marketers alike, is how we can approach content creation in a way that present us with a higher probability of achieving success. One way to approach this, is to study previous academic research and incorporate the frameworks they have developed. If content ideas are then compared against these, it should be possible to weed out the ideas that don’t fit the correct criterion.