BrightonSEO Breakdown - Part 2

by SEO Team on Thursday 11 October 2018

Fundamentals - Daniela Lui

How to Optimize for Visual Search by Clark Boyd

Usually, when performing a search, the input is text. However, in the case of visual search - as its name suggests - the input is a picture. Visual search turns a smartphone camera into a visual discovery tool, allowing consumers to scan an object or landscape and get relevant results in return. Some of the main visual search providers are Pinterest, Google, Bing, Snapchat, Amazon and ASOS.

According to Pinterest’s Amy Dziewiontkoski, shopping has always been visual - we’ve just been taught to do the opposite online. Visual search creates a new space for image-driven, inspiration-based connections. Consumers are more likely to engage with brands through visual media. Google and Amazon (as well as other major companies), having realised this, have adapted their strategies accordingly: Visual search will be integrated with Google Express, Snapchat’s Snap Store, and already works with Pinterest’s Lens, where over 600,000,000 visual searches are performed each month.

To optimise for visual search you should, first of all, upload your product inventory to your website and social media profiles and create an image XML sitemap. Then, research trends and concepts. Pinterest Lens releases their top trending searches, so it is useful to have a look there at some point during this stage. Once your research is complete, map keywords to images. Google will prioritise images that focus on a particular product type or that appear as a complete look and of course, are from authoritative sites. Avoid using stock photography unless absolutely necessary and, in that case, edit them to make them unique. Finally, structured data is essential to help search engines understand your content. Finally always mark-up price, availability, image and product name.

Page Speed - Olivia Stone

Winning with Mobile Page Speed - Aleh Baryservich

Does a fast web page mean a quick win for rankings? Not necessarily. With the Google Mobile PageSpeed update rolled on 9th July 2018, every brand was eager to ensure their mobile pages were as fast as physically possible. However, Aleh Baryservich proved why fast page speed alone will not secure the results that every SEO is hoping for.

It’s no secret that Google is OBSESSED with mobile page speed. With the average mobile page taking 15.3 to load, but more than 50% of users wanting a page to load in less than 2 seconds, Google’s Mobile PageSpeed update allowed the slowest websites to be sifted out of the bulk. However, in order to see the results that everyone is looking for, Aleh Baryservich advises that we follow the RAIL Performance model, which prioritises optimisation over page speed:

Response – it must be completed within 100ms

Animation – it must be visually smooth (60fps)

Idle – what happens in the background (50ms blocks)

Load – first meaningful point should be completed within less than 1 second and the time to be interactive – less than 5 seconds

But what speed metrics are the most important to consider in this optimisation process? The recent development of the Google PageSpeed Insights tool gives an indication of this. The tool now places an emphasis on speed AND optimisation, with additional features that measure the FCP (First Contentful Paint – the first visual response from the page), and the DCL (DOM Content Load – the document has been parsed and loaded). It also provides two scores out of hundred: one for page speed alone and another for optimisation.

How do we improve this optimisation score? Countless ways. Improving server response time, avoiding landing page redirects, minifying resources, prioritising above-the-fold content are only a few of the ways to boost your optimisation score from red to green.

So, yes page speed is important for rankings. But, for real results, we should prioritise the optimisation of our web pages, which will indirectly improve page speed anyway. Ultimately, if we follow the RAIL Performance model, we’re on the track to success and better rankings.

Personnel Development – Andrew Hoy

Why Failure Should Be An Option For Us All -  Arianne Donoghue

The talk initially focuses on how women are less likely to apply for roles within tech industries. This talk raises an issue on an important topic within our industry that often get brushed under the carpet.  Even though Arianne covers the subject of there being a lack of women applying for tech roles, her teachings can be applied to both genders.

Figures have shown that most women only apply to a job when they meet 100% of the criteria whereas most men will apply when they only meet 60%. This is driven from young girls developing a fixed mindset from a young age, where even though they have the same level of talent lack confidence when compared to boys. This is replicated in later life where 83% startups have male founders and only seven CEOs in the current FTSE 100 are women. To put this in perspective more people in the FTSE 100 are called Dave than are female.

To combat this, Arianne recommends that we all develop a confidence to fail. Creating a mindset of ‘moving fast and breaking things’ is beneficial for personal development as it sets a precedent for you to soak as much information as possible while not being afraid to fail.  To do this we have to be bold and reframe in our minds what it means to fail by understanding how important failure is to the learning process. She creates a seven-step plan to encourage you to fulfil your potential by building up your confidence regardless of your gender.

1.)   Plan – plan early on what industry you would like to work in and understand how to get there.

2.)    Know your shit – do your homework and know about the industry as possible.

3.)  Don’t be scared – Apply to jobs or go for positions even though you may feel that you don’t meet 100% of the criteria.

4.)  Be positive – Failing and receiving setbacks can be positive. Frame it as a learning experience.

5.)   Keep a log – Keep track of what you have done/ learnt and not what is required to move forward.

6.)    Be bold – Always be ready for new and exciting challenges.

7.) Be visible – Make yourself heard.

Voice Search – Rita Santos

“Optimising customer experience in the age of conversational AI”- Purna Virji

Purna Virji opened her talk about marketing in the AI-era of search at BrightonSEO by explaining that when facing two different ways to reach the same result, naturally the human being will always choose the easiest and fastest way. Apparently voice search would be the easiest and fastest way to do a big part of the searches we do in the internet every day, but that is still not 100% true.

Voice search is not perfect yet. We all have stories about how Siri misunderstood what we asked her about or what we asked her to do, and this could be the reason why voice search is not as widely used as it potentially could be.

On one hand, when perfected, voice search could be used almost effortlessly. On the other hand, voice search is not competing against a perfect system. So, brands that make their voice interface with customers easy to use, effortless and problem free will have a huge advantage against their competitors when aiming for customer loyalty and sales growth.

In order to make this adaptation to voice search as effective as possible, Purna Virji talked us through the 4Cs of customer experience, or the four principles of conversational design:

Clarity, Compassion, Character and Correction.

Clarity - when communicating to a client through a voice interface it is crucial to be clear when presenting different choices. Two good options would be to name all the options and then ask the customer to choose one, or to ask the question and then give the customer an example of a possible answer.

Here are two examples of the situations stated above:

  • “The doughnut comes in three flavours: chocolate, strawberry and vanilla. Which flavour would you like?”


  • Please name a stock you want to look up. For example, you can say Goblex Corporation”

Character – it is important that the bot users will talk to is perceived as nice, polite and friendly, however, it is very important that users can clearly understand they are speaking to a bot and not a human being. One of the reasons why this is important is because we are more forgiving of mistakes if we know we are talking to a bot than if we think we are talking to a human.

The personality of the bot should be chosen according to its purpose. For example, maybe an insurance company should have a bot with a professional personality, but a dance school can have a friendlier one.

Compassion – a good example for this is small talk. It is something that humans like a lot, but it is also an area that bots struggle with. A suggestion to improve this would be to build various small talk scenarios and make sure we incorporate them in our bot’s vocabulary, so it can maintain a nice conversation with our users and avoid saying “Sorry, I don’t understand.”.

Correction – sometimes bots misunderstand what we say, and in those situations it is important that they know how to correct the mistake and maintain the dialog with the user. For example, if we search for a flight from Newark and the bot understands New York, instead of simply saying “Sorry”, the bot can correct its mistake and immediately proceed to the next question, as the next example shows.

  • “Where are you flying from?”

  • “Newark.”

  • “Flying from New York. Which date?”

  • “No, Newark.”

  • “Got it. Flying from Newark. Which date?”

When planning how bots will interact with our users it is good to keep in mind the following topics:

  • Focus on the need for clarity;

  • Choose a suitable character for our bot according to our business area;

  • Make our bot empathetic;

  • Prepare our bot to make corrections when necessary and to keep the conversation flowing.

If you would like to catch up on any of our other SEO articles, you can find them here.