Solvemate is a chatbot builder that “enables meaningful conversations”. With their platform, you can create chatbots on websites, Facebook Messenger and WhatsApp. But should you use Solvemate to create your chatbots?
Let’s dive in!
What is Solvemate?
Solvemate is a “customer service automation platform powered by smart conversational AI”. They mainly focus on companies with a high volume of repetitive support questions and at least 15 customer service agents.
Before we deep dive into the various elements of Solvemate, it is important to understand how their software works. Solvemate uses a combination of dynamic decision trees and natural language processing to create chatbots.
With “normal” decision trees, you ask several pre-determined questions to the user and based on the answers of the user, the chatbot will provide certain messages. However, with normal decision trees, the questions that you ask are always the same (just as the message of the chatbot).
With dynamic decision trees, this is different. Instead of always asking the same questions to the user, you ask only the questions that you think are necessary. And this is what Solvemate does.
With Solvemate, you only have to provide certain questions into their system, and Solvemate will decide which questions are relevant to the user:
This comes with its pros and cons (as you will see in the review). On the one hand, if you will deal with thousands of questions every day (and the algorithm works well), the system can choose only the necessary questions for the user and ask them, which is really efficient.
On the other hand, this system heavily complicates the interface of Solvemate and it is not easy to do simple things, such as asking questions without needing their help.
When you start for the first time with Solvemate, you will get through a step-by-step guide, that will help you get started with the Solvemate platform (which is pretty useful):
When you have gone through this guide, you can create your chatbot flows (called Solutions in Solvemate). Each solution starts with an identifying question, which is a yes/no question that specifically asks whether the solution is what the user is looking for. This system is required for each solution, which can be a pro and a con (depending on what you want to do).
Therefore, each chatbot flows generally consists of 3 elements:
- The user asks something to the chatbot
- The chatbot asks if this is the solution the user is looking for (the identifying question)
- If the user replies with “Yes”, the chatbot provides an answer
And this is how you can set that up:
However, it couldn’t get it to work. To be honest, the interface of Solvemate is a complete puzzle to me and I don’t think it is easy to set up your chatbot.
First of all, I do not understand why each chatbot flow needs to consist of an identifying question, because there are many use cases in which you don’t need an identifying question. Secondly, when you’ve determined the identifying question, I didn’t find a way how you can create an answer for that question. Also, for me, it is unclear how the chatbot flows relate to each other and how you can activate different chatbot flows.
After speaking with a Solvemate representative, he told me that for simple things such as asking questions and creating questions, instruction by Solvemate is almost always necessary. Also, in many cases, Solvemate first builds the foundation of the chatbot and provides training, before the customer will start with using the chatbot.
- Visual Flow Builder
- Test chatbot
- Interface easy to use?
- Easy to setup a chatbot?
The chatbot elements of Solvemate are pretty minimal.
With Solvemate, it is only possible to use images, emojis, quick replies and buttons.
- Gallery (Horizontal list)
- Vertical list
- Persistent Menu
- Quick Replies
- Free text input
- Location sharing
- Email validation
- Phone number validation
- Date validation
In Solvemate, it is possible to create custom fields (called Variables).
However, I did not succeed in asking a question to the user and saving the answer in a custom field. For this, you probably need some training by Solvemate.
Moreover, it is not possible to create tags or user segments in Solvemate.
- Custom fields
- User segments
In Solvemate, you cannot send broadcasts and sequences. Essentially, every conversation in Solvemate is initiated by a question of the user.
Moreover, it is possible to use conditions, but you need help from Solvemate to do this. The Solvemate representative told me that it is possible to personalize the chatbot experience with the following variables:
But again, I have never tested this out, because you need help from Solvemate to implement this.
Also, it is possible to make a chatbot in any language, but it is not possible to handle multiple languages at the same time.
- Multiple languages
- Easy to translate
AI & NLP
With Solvemate, it seems that the chatbot can recognize keywords, but I never got this to work.
This is also true for intents, it seems that the chatbot can understand the intent of the user, but I also couldn’t get this to work either.
I think for both of these features, you also need the help of Solvemate.
- Dialogflow integration
With Solvemate, you can put your chatbot on:
- Your website
- Facebook Messenger
- Your app
Solvemate provides direct integrations with various live chat applications:
- Focalscope Chat
But these are almost all the integrations they provide. They don’t provide integrations with Zapier, Integromat and Google Sheets, and it is not possible to create your own webhooks.
- Google Sheets
With Solvemate, you can put an advanced native widget on your website and create a landing page.
The landing page of Solvemate is pretty simple, it is just a blank page with a chat widget on the right bottom:
The native widget is a little more advanced. You can use the native widget as a chat widget (just as on the landing page), but you can also use it embedded on a website page:
Also, with both the native widget and the landing page, you can customize the styling of the widget:
What is a little bit of a drawback with a chatbot builder so focused on websites, is that it is not possible to show different greetings per page on your website.
- Facebook website widget
- Native website widget
- Landing pages
- Facebook Ads
- Facebook comments auto-reply
- Button creator
- Custom URLs for your chatbot
Solvemate provides a 30-day free trial on their platform. This gives you plenty of time to test out all the features and decide whether Solvemate is the right fit for you.
The first plan of Solvemate will cost you €349 per month and comes with a chatbot for web and mobile, proven bot templates for key industries (I haven’t seen these), a chat widget, built-in AI, email & phone handover and analytics.
If these features are not enough, you can also go with one of their more expensive plans:
Moreover, if you are choosing Solvemate, I would always recommend having some onboarding (this is also what Solvemate themselves recommends). Their onboarding plans start from €500 and range from providing training, to Solvemate creating the chatbot for you:
In general, I think these prices are pretty expensive for the features they provide.
Also, as you can read above, I couldn’t get the chatbot to work, which was a huge drawback.
- Free trial
- Free plan
- Price per month349
- Monthly pricing available?
- Money back guarantee
Solvemate provides a dashboard with the number of started conversations, number of solutions, the self-service rate, verification rate and time-to-solution:
Moreover, the dashboard shows the conversion funnel and the customer satisfaction score:
I think these analytics shows some basic insights in your chatbot, but don’t provide enough information to improve the answers of your chatbot. Also, Solvemate doesn’t show the questions your chatbot didn’t know an answer to, which is a disadvantage.
- User input
- User retention
- Sent messages
- Open rate
- Click rate
- Conversion rate
- Fall back rate
- Volunteer users
- Custom dashboards/reports
Unfortunately, Solvemate doesn’t provide any direct eCommerce integrations.
- Buy directly within the chatbot
- Cart abandonment
Templates & Cloning
Solvemate provides no templates that you can use. Also, it is not possible to create templates yourself and share these with others.
Moreover, you cannot copy your chatbot flows and chatbot elements.
- Make template
- Share template
- Duplicate chatbot flows
- Duplicate chatbot elements
Solvemate doesn’t provide any optimization, such as A/B testing.
- A/B Testing
Solvemate provides some tutorials on documentation on how you can use their platform.
But in my opinion, the documentation is mostly focused on developers and not so much on explaining how their platform works. As I already mentioned before, there are some things that I couldn’t get to work and their documentation also didn’t point me in any direction.
Moreover, they provide customer support through email.
- Customer support
So, should you buy Solvemate to create your chatbots?
It depends on the type of company you are (and I don’t say this often).
I think I would never recommend Solvemate to a small or medium-sized company. Often these types of companies want to do things (partially) themselves, and that is not possible with Solvemate (you are almost always reliant on Solvemate). Also, I think their pricing plans are too expensive to make a decent return on investment.
But, if you are a large enterprise company that needs to handle thousands of customer service questions every day, Solvemate can be interesting. Using the dynamic decision tree algorithm (assuming that it works well), it could save you tons of hours creating chatbot flows, because the algorithm decides the most optimal questions for the user.
That being said, with Solvemate, you also need to think about your reliance on the company. Solvemate is definitely not a do-it-yourself platform and you probably need some onboarding (and them creating the chatbot for you) to get started. Also, I think Solvemate is primarily useful for customer service chatbots, and not for chatbots focused on marketing.