Automation

In this blog we will cover 5 topics on Ecommerce automation and discuss what improvements can be made when you implement strategies from Retail:

1. Automated marketing and ads
2. Automated personalized content
3. Automated follow up
4. Automated customer understanding
5. Automated interaction and advising

This is part 6 of my series on “5 Things Ecommerce can learn from Retail”. Ecommerce has had it easy till now. It was sailing with the wind and almost all companies were successful. But competition is getting tougher. Ecommerce has become like a normal trade, where some have success and others fail. It has become increasingly important to learn and improve. My vision is that retail selling strategies have a lot to offer.

The first blog on this topic covered customer education . The second one was on building customer value. The third blog was on removing obstacles. The fourth blog was on persuasion. The 5th and last blog was on the importance of showing activity to customers.

I intended to write 5 articles about my vision that Ecommerce can learn a lot from retail. Now that that objective is accomplished, I still have ideas left. There is at least one that I want to share with you, so I decided to add that to the series. This blog is about automation, a key topic for Ecommerce. To heat up the discussion right away, let’s start with a statement:

Automate as much as possible, but not more.

I have to admit I more or less copied this quote from Albert Einstein. His quote is:  “Make everything as simple as possible, but not simpler”.

1. Automated marketing and ads

What is easiest to automate is advertisement. People actually expect advertisement to be automated / programmed. Advertisement can be characterized as the confrontation with a certain product without having taken action to look for that product.

Automated advertisement however has its limits. We see more and more “ad avoiding” customers. A typical example is that “ad-haters” troubled the video service Hulu so much, it decided to offer an ad-free service as an additional option.

The modern version of advertisement is content management. It is more acceptable, because the company actually gives something to the reader. But as more and more companies started to use this strategy as an alternative for advertisement, the overflow of content offered is overwhelming. People start avoiding content as well.

What’s liked best is photo and video. When offered to a customer, video also looks like a photo. Instagram and You Tube are rising in popularity as marketing media. In my vision the main reason is that a picture is the best acceptable way to communicate an unrequested message. Because of the holistic experience photo is the best medium to be catchy, fun, unique in a split second.

I like to compare this to the ads in magazines. People love the flashy magazines for their beautiful ads. The value is in the photo itself, not in the advertised product. That’s why nobody ever complained about the huge portion of advertisement in these magazines.

2. Automated personalized content

Personalization is hot in Ecommerce. There are many tools to produce content that matches the profile of the customer. Examples are:

• presenting similar products
• presenting complimentary products
• presenting “other people also looked at”
• adjusted sorting based on customer journey
• less frequently used: special deal offers related to the customer journey

These are Ecommerce improvements that help the customer navigate through the site. It’s like a shop assistant showing you a cardigan that goes really well with the trousers you are trying on. This is really a win-win approach. It safes time for the consumer and the shop increases its conversion.

What can still be improved is collecting the feedback of the customer to the proposed products. The tools collect the click, but it would be great if the customer could actually reply something like “I like that, but it’s too expensive”. The shop assistant goes through a very steep learning curve with each individual customer. It would be great if that would be possible on internet as well.

3. Automated follow up

One of the best conversion boosters in Ecommerce is automated mail marketing. It is also widely used in retail. There are many strategies for this and all have a great impact on sales. To mention some:

• Advice on how to use the product
• Products that go well with the product just bought
• Renewal of the product, at the end of the normal use period
• Hot deals your customer might not want to miss
• News and innovation about product, the assortment or the company
• Congratulation at customer’s birthday.

All of these strategies build on the well known fact that it is much easier to persuade an existing customer to buy a next product from you than to persuade a new customer to buy a product from you for the very first time. It is said to be a factor 6 easier. The reason for difference is TRUST. The existing customer has experienced your service and taking that your service and product were good, he will not have to cross the trust barrier again.

We see two types of behavior here. Many consumers like these messages. This leads to the high conversion that results from these mails. But we also see ad-avoiders, who have opened free mail accounts specifically to receive these mails and trash them. Most customers will do both. Their choice of strategy depends on the relation they feel towards the company. Here Ecommerce can learn something from retail.

Ecommerce is very quick with asking for mail addresses. But when the relation is not yet build on trust, you will get many fakes. Retails spends more personal quality time with the individual customer, building the trusted relationship.  Now the customer is more likely to give his real email address and marketing mails will be much more effective. The advice would be to accompany the request with more and more personal attention, not to reduce the request for email addresses.

A second form of automated follow up is the ad banner, Google Adsense, Facebook targeting etc that are triggered by tracking the customer continously during his internet presence. Much is known about the income these advertisement generate for companies like Google and Facebook. There is also quite some information about the income it generate for the banner/advertisement hosting sites. Much less is known about the conversion they deliver to the company that pays for the ad. Of course, every situation is different, so it isn’t easy to compare. The most objective available data we found showed an average Click Through Rate of around 0.1%. That still doesn’t say anything about conversion. We have other research from SeeWhy that shows that this kind of traffic is 15 times less effective than mail traffic. Might we be looking at the cloth of the Emperor? If anyone can help me to objectively measured conversion rates of these forms of advertisement, I would be most obliged.

Might we be looking at the cloth of the Emperor?

What I do read a lot about is about people avoiding or blocking Adsense and banners. Adblockers are very popular in the App stores. Up to 35% of all ads are blocked. Google reports that less than 45% of ads are actually viewable.

My hypotheses however is that, on top of the above, most people have a trained blind eye for these banners and ads and just ignore them all together. There has been a research on the effectivity of advertisement. They let people read a newspaper with ads in between. One mayor expensive ad was tweaked. The advertised product and the brand of the product clearly didn’t match. None of the participants noticed.

4. Automated customer understanding

Now we are getting to a more difficult topic: understanding the customer. There are many companies that try to innovate in this area. Conceptually there is one problem. One customer account mostly represents one person, but not always. Possibly the account is used by the whole family, leading to different behavior. Similarly one person has different roles in his life. Employee or employer, husband, father, tennis team member, care taker for his elders, etc. Each role will also lead to different behavior. Without interaction it is pretty much impossible to understand an individual. What is easier to understand is the trend in behavior by the mass of consumers. That is where the tools in Ecommerce focus on.

When you look at this from customer perspective, you will feel little appreciation for being treated as average. Every customer likes to be treated as a unique person. Mismatching are quickly causing irritation. In this area retail clearly still has an advantage.

Retail invests in focus on the individual. A shop assistant will ask what the customer wants and what he wants to achieve with it. In this way he can understand the customer as a total person. The weak point in retail is that it is much harder to get the statistics. In Ecommerce every movement is measured and chances of follow up actions can be calculated. In retail that hardly has an equivalent.

The difference between Ecommerce and Retail can be summarized as

Ecommerce works from statistics of actions to predict the needs of one customer

versus

Retail works from their understanding of individual customer needs and tries to predict the general trends in behavior.

Ecommerce has the potential to combine both approaches, when it would be possible to interact in an open environment with the customer. This is different from the customer service chat. It should be experienced as an open interaction where the customer can act and interact without barriers.

5. Automated Interaction & advising

Ecommerce has many forms of automated interaction. You can distinguish two types, static and dynamic. Examples of the static type of interaction are the frequent asked questions (FAQ). By predicting the question, Ecommerce can give the answer on a static page. This works very well and save a lot of costs.

Reviews might be a form of personal interaction or also a form of static interaction. That all depends on the response times. When reviews are not answered or with large delay, they are a form of static interaction. The customer still gets a sense of past interaction, but he will not expect to have any interaction himself from the company. When the response follows quickly, they become a form of personal interaction.

Ecommerce also has the automated assistant. These tools allow the customer to formulate a question. The tool will then try to understand the question, may propose some alternative questions that might cover the original question and then supply an answer. Of course this is also a great cost saver. The success of these tools vary much with the complexity of the question. These tools are in a squeezed position. When the question is very simple, you could have solved it with a well organized FAQ. when the question is complex the chance of providing a satisfactory answer is very limited. Basically these tools are FAQ with a search index.

Ecommerce has since a number of years the customer service chat plug-ins. These brought a great improvement. This is a form on personal interaction that is oftentimes very effective. The customer is served personally by the service center. Mostly the name of the shop assistant is published, which gives a more personal image to the service center. What is automated is the push of the box. I personally find this highly irritating, but apparently other people have a hard time finding the plug-in.

In retail personal interaction has a much bigger impact. There are a number of elements that contribute to this effect:

• the shop assistant is physically visible
• eye contact between customer and shop assistant communicate the (absence of) need for assistance.
• while communicating, the customer perceives the focus of the assistant on him.
• Non verbal communication is added to the pallet of interaction
• when returning more frequently, the shop assistant and customer will recognize each other. This creates engagement.
• customers can oversee / overhear the interaction of the shop assistant with other customers and thus experience the service mentality of the company. This also creates livelihood.

If Ecommerce could implement some of these elements, the impact of interaction could further increase engagement and conversion. Going back to the statement at the start, we should automate everything possible, but not more than the possible. In my view customer interaction can not be automated adequately. At least not at this moment in history. There might be a time in the future when computers/robots will understand humans and can really interact, though I don’t like the idea of it.

What do you think?

With courtesy to Flickre.com for the photos