We have discussed marketing automation on the KoMarketing blog and industry news section several times, including speaking to how important it is, how to create content around it, and why so many marketers still use it as it improves performance.
Even though marketing automation can be an effective way to reach more customers faster, there are some times where it doesn’t make sense and can actually cause a poor customer experience. This is usually because the automated response isn’t what the customer is looking for, leading them to become even more frustrated, confused, or even angry.
In the below sections, we’ll discuss when B2B marketing automation is NOT a good idea, and what you should be doing instead.
Customer service has transitioned to helping customers on platforms that go beyond the phone. Social media, and Facebook in particular, is becoming one of the more popular ways consumers are looking to reach out to businesses, no matter the industry.
According to Pew Research Center, 68 percent of all adults in the United States use Facebook. This is an overwhelming majority and makes Facebook a huge leader against all other main social networks (for perspective, only 25 percent use LinkedIn). Ever since Facebook debuted its new Messenger platform and review options for business pages, customers have been using the social network to connect to businesses with complaints, questions, or praise.
Facebook allows businesses to automate some of the messaging process with consumers by allowing Instant Replies and bots to try to solve user queries automatically. The thought behind this is that if more customer issues could be solved with AI, then businesses would need less employees to answer common questions.
While this works for many types of businesses, B2B customers typically expect a higher level of support than an automated message. So instead of setting up an automated reply on Facebook messenger or email, consider setting up the proper resources to allow your support team to answer questions quicker, but with a human touch. Online customer support platforms like Zendesk and Freshdesk present customer inquiries in a intuitive format that makes it easier for employees to respond. This can give customers the human touch they are looking for, and helps employees answer more queries in less time, thus allowing them to help more people.
If you have any type of blog on your website, you’ve probably been inundated with canned emails asking for links or guest post opportunities so they can submit content. While it does require a human to actually send the emails (if they haven’t put you on a mailing list without your permission to further automate it), the text itself is very automated. Many times the person will plug in your website’s name and link to their prepared email before sending. This allows him or her to send hundreds of emails per day.
While this method of reaching out to websites and blogs for exposure isn’t necessarily a bad thing, it needs to be done correctly and with special attention paid to each individual blog owner. Reaching hundreds of website owners per day is useless if none of them respond because their name is spelled wrong, your content doesn’t match their website topic, or they can obviously tell your email is the same one that’s sent to everyone on your target list.
When it comes to targeting websites and influencers for campaigns or guest blogs, it should be quality over quantity. Creating a great relationship with one website that is visited by your target audience is better than partnerships with less reputable websites that really don’t get much traffic.
Trying to use canned, unspecific messages to build relationships, whether it’s outreach or sales, is going to eventually come back to haunt you. At best, you’ll get ignored, but at worst, you’ll make people mad and they won’t want anything to do with you.
Take this actual example I received for my business website, which does digital marketing:
I ignored the email but got a follow-up email from him a few days later. It made me mad because it was a canned email, my website has nothing to do with personal finance, and he used my old business name, which I haven’t used in almost a year (and is no longer mentioned anywhere on my website except to mention that it’s now Six Stories). This resulted in a snippy reply back from me:
While the people you’re targeting might not be as direct and short as I was to “Tom,” the point is, if people aren’t responding to your emails, they aren’t interested. Instead of trying to automate the process of building relationships, try hunkering down and just getting to know people genuinely first, through an introduction or social media, before asking them for something that doesn’t benefit them or their business. Most people are unlikely to respond to a request for something if they don’t know the person or see an immediate benefit.
Ad creation is another aspect of B2B marketing that necessitates a human touch. This point may be contentious to some, but when it comes to B2B, automated ad creation likely may be a waste a budget and time. Facebook, Google, and other platforms offer dynamic ad creation, which basically takes the keyword the user is searching for and inserts it into an ad that is pre-written to fit any query in a specific category. While it works perfectly for many consumer queries, like this example for “nordic walking poles” from Amazon:
When you click on the ad, it takes you to the product search page for “nordic walking poles,” which is presumably what a customer wants with that type of query.
However, for some user queries, dynamic ads just don’t work. Either the landing page doesn’t give the user the information they are looking for, or it makes them work harder just to find what they want. Take this example for “buy copier” which brings up an ad for OfficeDepot:
When you click on the link, it takes you to a page about all-in-one printers that are also copiers:
We can presume that when someone is searching for “buy copier” they want a copy machine, not an all-in-one printer. This is a good example of a potential B2B query that has completely missed the mark with dynamic ad insertion.
While dynamic ads can help save time for some websites (like e-commerce sites that have matching product search result pages), for others, they just don’t work. Depending on your industry, you may reach more customers by creating more ads manually with a broader keyword list. Even if your ads don’t match the user’s search query exactly, you likely have a higher chance for a click if it better matches what they are really looking for. This example of “virtual assistant services” ads showcases that:
All of these ads have a higher likelihood of matching a user’s query and intent than a dynamic ad may have, depending on ad copy and landing page.
When it comes to data collection, it’s perfectly OK (and encouraged!) to automate the process. Sending automated reports is a great way to save time while ensuring that data continues to be sent to the right people. But where this often falls apart is the “review” part of the process.
Automated reports can save us time (QuillEngage is a great example of an automated report that explains data in layman terms), but what is important about that data is what we do with it after the fact. There are some platforms that allow you to automate decisions based on data, such as those mentioned in this AI tool roundup by DailySEOBlog. While AI can certainly help us as marketers get menial tasks done more quickly (e.g. transfer content across platforms, or email reports), some aspects of marketing just shouldn’t be left to AI.
For instance, let’s say a client is seeing a drop in time on site for their top target country. Looking at that data alone, AI would likely choose from a list of potential causes, such as site speed, content (e.g. is it optimized for country-specific keywords), or site errors. However, if the AI makes a change based on this list and nothing changes, we still have the same problem with no end in sight. An experienced marketer’s eye can look at the campaigns run, what’s happening in the country, and feedback from customers to make more educated decisions about what to change on a website.
Another example would be those automated “SEO audits” that SEO tool platforms offer. These check for a variety of issues your site may have, such as missing title attributes or an incorrect sitemap. What these won’t tell you is how your content, design, and website as a whole looks like to a potential customer. You may get 100% on an SEO audit from an AI perspective, but the website content may be boring and the navigation clunky. These types of issues are impossible to diagnose without a marketer manually reviewing what could be wrong.
Automation That Hasn’t Been Tested Before
This should go without saying, but I’m still amazed at the number of emails, bot messages, and marketing campaigns I see that are clearly automated. Before launching an automation sequence, whether it’s a bot or an email drip campaign, make sure it’s reviewed and tested by real people.
For example, a clothing website I enjoy automatically connected me with their Facebook bot since I log into their site with my Facebook account. This bot automatically tells me when I’ve placed an order, when it’s shipped, etc. While I appreciate the sentiment behind this (e.g. the customer wants to know the status of their order from purchase to delivery), it doesn’t mean I want to be receiving automated messages I didn’t manually opt into. In addition to order updates, the website bot was also sending me promotional campaigns.
While this could have been successful for the company overall (and I’m sure it varies by industry), I wonder if they did any sort of A/B testing or asked for feedback from actual customers before rolling it out to all customers.
When you are automating any part of your process, go through it like an actual customer. Place a real order or complete a lead generation form to see what the process is like for an actual customer. Looking at it from their perspective, you may notice annoyances or issues in the process that actually makes it more frustrating, not easier.
Once you think you’ve ironed out the process, ask a few real customers to try it out and get their responses. If it’s something on your website, you can use a testing service like UserTesting.com or even recruit testers on sites like Fiverr or ask local organizations if anyone is willing to try it out for a free product. You may be amazed at the feedback you receive when you simply ask.
After all, I have never told that clothing company I disliked the bot, I simply put it on mute. They can likely tell that I’m not interested in promotions because I’ve never clicked on a bot link, but otherwise, they may think I “don’t mind” the service when, in actuality, it has decreased my order frequency just because they never asked my permission in the first place.
Automation is supposed to make our work as marketers easier, but is it really worth it if it doesn’t give the customer a better experience? Before implementing any automated systems into your marketing, make sure it makes customer experience quicker, easier, and better, instead of setting it up simply because are able to.
Screenshots taken February 2018. Featured image via Pixabay.