Blast campaigns, personalization vs. scale and a product-centric focus instead of a customer-centric approach. These are the three headaches that plague so many online marketers and cause them to lose the battle for retention, money, and several other areas where they could win. Add to these the misguided focus on analytics and misinformed KPI settings and you have a real loss-stew cooking on your stove.
This is according to Tanya Szwarcbard, Director of Professional Services at Optimove, a smart marketing-automation software that specializes in predictive analytics and customer-level prism (and the parent company of PostFunnel). But Tanya doesn’t only point out the sickness, she also offers a treatment. We sat down for a talk about what Professional Services really do, the main problems marketers suffer from, what metrics they’re focusing on but shouldn’t, and lastly, what marketers need to do to be better at their jobs.
The Professional Services (PS) team is an independent business unit comprised of around 20 employees and managed on a project-level matrix manner. They either help clients make the optimal usage of the product, thus maximizing their business potential, or they provide additional, complementary services that do not exist in the product itself.
Tanya Szwarcbard, Director of Strategic Services at Optimove
You are working with companies in different segments – e-commerce, gaming etc. – what are the most prominent pain-points you can identify in their marketing operations?
“The first and most prominent pattern is the tendency to send blast email-campaigns. Even within established and successful companies you won’t be able to avoid it, because many of them would say something like, ‘with all due respect to personalization, there are some current events I cannot avoid, like the World Cup or Valentine’s day.’ But if someone just signed up to your service, he should first get a personalized onboarding email and only then additional materials that are more generic. Many companies gradually start to overcome this tendency and realize there are solutions out there, like dynamic templates, that allow them to separate the different customer types from other funnels, but most of them are still trapped in old habits.
A second pain-point has to do with the conflict between personalization vs. scale – how much can you personalize your marketing efforts while still keeping them manageable? For example, you create only four campaign segments, but slightly personalize the messaging in each segment, so, on the whole, you can reach a much higher number of individualized campaigns.
Here’s an example from the world of gaming: Let’s say you have two customers in the same lifecycle stage, but one has just lost a game while the other won the game. Normally, you’d send both of them an automated message urging them to play more, but it proves to be so much more efficient to send the former a message with something like ‘don’t despair, keep playing’ and the latter ‘way to go, now keep up the good run.’ And our system makes this process so much easier because it automates this entire process.
Lastly, but very importantly, I believe many companies are not as customer-centric as they perceive themselves. I’ll explain this using an example: There’s a very large company that has worked with our PS team since their first day with our product. As soon as we began working together, they sent us their organizational chart to review and help us learn their structure. What we noticed, was that they have many products (casino, sports, fantasy, etc.), and the company’s structure was built around each product. It’s what we call product-centric structure: The client bases, marketing methods, and campaigns were separated and the retention strategy differed from each product. The end result was a very unhealthy state, where they might treat one customer as churned (or about to churn) when in real life, that customer is very much engaged with the company and only moved to another product they offered. What we did was recommend changing this structure, so that each customer is examined on the whole, and not through the product’s perspective, i.e.: create a CRM team responsible for all the customers of the company, and within that team, have team members who “represent” each of the company’s products. This proved to be incredibly cost-effective for the company, because now, a customer who churned is a customer who was not active on any of the company’s products, and retention can be much more efficient.”
Are there any cases where it is not recommended for a company to be customer-centric?
“Not in the world of B2C, no. At the end of the day, the number of customers and their level of satisfaction from your product will determine your company’s future. What I can say, is that there are situations where being customer-centric is simply more difficult to achieve. For example, when you’re a multi-national conglomerate. Japanese customers are inherently different than Argentinian ones, and you can’t have a CRM team attending to all these customers unless you have regional representatives.”
Let’s zoom into the world of CRM. What is the main obstacle?
“I think it’s very evident that the market understands how it is vital to be customer-centric and try to work around it, despite the obstacles. One major obstacle is when your company has several geographic bases, like some of our clients. It’s not at all easy to remain customer-centric when one major product is operated in one country and the other from another country. In fact, this is such a sticky situation, that it causes some companies to withdraw from their customer-centric approach back into the product-centrism. Even from an operational perspective, I can totally understand their approach, but from a marketing perspective I’m against it.”
Performance analysis is always critical. What is the #1 metric companies should focus on and what are the metrics they shouldn’t?
“I can tell you that Optimove does not measure success using ‘traditional’ metrics like open and click rates, but rather by measuring actual activity, meaning – money that goes into that company’s pockets thanks to a specific campaign. If a customer received an email urging him to shop at Uniqlo and didn’t even open it, it doesn’t mean it won’t send him to Uniqlo’s online store and order that jacket he wanted to buy some time ago. The subject line of the email is enough to get him to do that.”
How can you relate that purchase of the jacket to your campaign if you’re not looking at whether the customer opened the email or not?
“I can measure this by using control groups and comparing them. Just like with any scientific experiment, you must use control groups. If a client opened your email on his mobile but did not download the photos, it would not be counted as open, so why focus on open and click rates? It’s not that these metrics should be ignored, but they’re not the main metrics one should use to measure a campaign’s success.
Another issue with analytics is that while many companies use good BI tools that provide very high-level insights, they don’t reach the level of a customer. These tools are good for the company’s management, but to really help the CRM teams understand – and act on – their clients’ behaviors, a product like Optimove is much more suitable, as it allows the understanding of customer behavior all the way down to the level of the single customer.
Lastly, many companies prefer the short-term revenue over long-term. That’s a problem because not many companies are willing to realize that in many cases, you lose money on new customers and start seeing revenues from them only in the mid-to-long term. It’s important to adjust the strategy to the lifecycle stage because I’d be much more willing to lose money on new customers, where it’s all about turning them active, whereas with active ones I’d be much less willing to lose money.”
Let’s move to setting KPIs – what are your recommendations?
“The most important thing is to understand is that KPI setting must differ from one segment to another within that company. You cannot demand similar results on different segments. For example, with new customers, the most important KPI should be how many customers went beyond the barrier of one purchase or bet. On the next level, you should measure and set your desired rate of customers that go from being one-timers to becoming active clients, or rather churn. Leave revenue KPIs aside and focus on the level of engagement – the revenues will come as a result. Lastly, companies should set KPIs that focus on expanding each customer’s usage of that company’s products. So if you’re an e-retailer, you’d want a beauty products customer to also start shopping for shoes.”
Optimove is a software company, focused on a technical product. Where does your team fit in?
“We give services that accompany and complete the services you get with the product, services that enhance your experience and benefits you receive from it. For example, our product allows the client to concentrate all his campaigns in one place, and we work with the client on the right strategy of how to manage these campaigns: Best practices; laying out the different campaign management approaches; where to start and how to proceed etc.
The idea is that we either build the customer’s strategy from scratch or if he already has one – we review it and submit recommendations. But there is one rule we always follow: Every recommendation we give, must be data-backed. Every member of the team has a strong background in data management and engineering, along with a deep understanding of data analysis methods and tools.”
We should state, though, that Optimove did not start off as a product company, but rather as a PS company. In other words, the company was built around your team, you were the bedrock.
“That’s true. Before we had the product, we specialized in data-based professional services for marketers. However, nowadays our team offers services that do not exist within the product, such as a module that predicts clients who have good prospects of becoming VIPs, using machine learning and data mining. We also created a statistical model to predict what the gaming industry calls ‘bonus abusers’, as well as predicting who will become an ‘irresponsible gambler’ etc. All these models do not exist in the product itself, but derive from the data we mine out of the product.”