cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Guides

Learn more about specific Flow features
Filter
Filter by Labels
Select any of the labels below to filter results.
Filters applied...
Sorted by:
Using Flow you can automate customer engagement in different ways. The following topics explain different scenarios and use cases: Implement Timer Integrate Zapier to the bot Increase web widget engagement Use Conditions  Set parameters using buttons Conversational Reviews Count sessions per Channel or Widget Entities instead of multiple intents Handle multiple user or customer messages Branch your chatbot Capture user input Change the language or brain of your bot   Clear the Web Widget   Duplicate the flows Create and manage multiple chatbots Create an Arabic speaking bot with right-to-left text alignment  Create buttons Export and duplicate the project Troubleshoot chatbot failures Improve your chatbot with String Templates Go to the middle of a flow Loop flows Reuse existing Intents Reset your Password Schedule and broadcast chatbot messages Design the Flow structure Set Business Hours Carousels for enriched UI Use existing data to create a context for your chatbot Use Quick replies Pause the bot Send and receive files or attachments The alternative option Measure feedback or performance Track chatbot links Implement WhatsApp Click to Chat Links option Accept payments via your chatbot Deflect phone calls to WhatsApp Implement Reply Delays Customer Engagement using Khoros and SMS text messaging Import contacts Test your WhatsApp chatbot Use Facebook Messenger's m.me link in your chatbot Multi-channel design
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
302
Timers   can improve the customer journey as well as the respondent rate when using correctly. When your chatbot hits the pre-defined time that is set by the timer it will trigger a follow-up. This follow-up can be sending a text, buttons, cards, and many more. That’s all up to you as a Flow-designer.   How do I create a Timer? To create a timer you can drag and drop the   Timer trigger   onto your canvas. After that, you can specify the following timer settings in the right pane. Interval This allows you to set the interval of your timer. You can set this to seconds, minutes, hours, and even days. Continue on reply When checking this box the bot will continue with the follow up action after the user replies. When unchecking, that follow up action will only be shown when the user is quiet for your pre-defined interval. Send typing This wil show the user an indicator that the bot is typing (if supported by your channel).   With these options, timers can add great value to your chatbot. Let’s explore 3 common options. 1. Complete registration A timer is great when you’re using a chatbot as a lead generator or to complete registration. In that case, the bot asks questions but when your customer doesn’t answer the question after an hour the timer will be triggered and you can follow up with an additional message. Instead of rephrasing the same question or   looping your flow , you can try to get answer to question in a slightly different way. Instead of an open question you could now ask a closed variant of the question and add quick replies if supported by your messaging channel.   2. Feedback request Measuring the performance of your bot is important. One of ways to measure that performance is by incorporating a   Feedback flow. In related article, that is just a simple question like “Was I able to help?”. To improve the customer experience, that “Was I able to help?” needs a delay. Otherwise, that question will be sent directly after sharing an answer. That’s a bit pushy. You might want to wait 5 seconds to give the user time to read and to give the option to ask another question without triggering that   Feedback flow   straightaway.     3. Handle waiting time Does your bot have a handoff to a live agent? In that case, you might want to add a timer after your handoff. That handoff will alert one of your agents to take over the conversation but if, for any reason, no agent assists within a few minutes your customer will probably leave. To let your customer know that it takes longer than expected, you can work with a timer after your handoff. In other words, we do a handoff and when there is no reply after 2 minutes we follow up with something like: “It takes longer than expected, would like to wait, or do you prefer creating a ticket?”. Now we’ve managed expectations and we offer 2 solutions! Or perhaps if you’re creative you can come up with a brand quiz while your customer is waiting for an agent to take over.    
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
215
Zapier is a service that allows you to integrate your bot to many other apps and tools. Connect your bot and check it out. Zapier    is a handy tool that allows you to integrate your chatbot with all sorts of different apps and tools. By combining your Flow bot with Zaps you can do more, without code. With a simple interface, you can easily extract data and add it to a Google Sheet, create tickets in CRM tools like Hubspot, Salesforce or Freshdesk and much more. Most chatbot cases can be divided into 3 categories. Therefore, we'll show you how to use Zapier for each of these cases: Customer Service Marketing & Sales Virtual Assistants 1. Customer Service - Create Tickets for CRM Chatbots in customer service are mostly used to automatically answer FAQ's on channels like Web, Messenger and WhatsApp. In some cases you might want to follow up with a human agent. Especially when we're looking into complex questions or questions where the bot is not trained for. Following up with an agent can be done in several ways. One of them is using a ticketing system. In this example, we will show how to create a ticket in Hubspot when the bot is not able to handle a question. Create your bot This is the place where you set parameters to extract like email and the message so you can follow up in Hubspot, Salesforce or other Customer Service Tools. Make sure to end with an action "Zapier Action". Configure Zapier Let's configure Zapier so that it creates a task in your favorite CRM when the Zapier Action is triggered within Flow. Read more about Zapier Configurations in our   Zapier Docs.     Create a ticket Wow that was fast. Your bot is now able to create tickets in Hubspot. If you're using SalesForce, Zendesk, Freshdesk or other customer service tools or ticketing system the process will be similar. 2. Marketing & Sales - Fill Google Spreadsheets Bots are also great for marketing & sales. You can combine a Facebook Advertisement and funnel customers straight into your chatbot for product recommendations or challenges. Also, what if your customer wants to contact you outside of business hours. Let's have a look at how we can use a bot to gather leads and fill a Google Spreadsheet. Let's break it down in 3 steps: Create Your bot This is the place where you set parameters to extract like name and email. Make sure to end with an action "Zapier Action". Configure Zapier Let's configure   Zapier   so that it creates a spreadsheet row every time that the Zapier Action is triggered within Flow. Read more about Zapier Configurations in our   Zapier Docs.     Generate leads Done, your bot should now fill rows within your spreadsheet.     3. Virtual Assistants - Add songs to a Spotify playlist By now, we've covered Customer Service and Marketing & Sales so let's continue with a Virtual Assistant. At Flow, we enjoy all kinds of music but not all of our team members have a Spotify account. What if they could just send a WhatsApp message with their favorite song and that song will automatically be added to the Flow playlist. Isn't that awesome! Again it starts with the creation of the chatbot. Create Your bot This time you extract the name of the song. Make sure to end with an action "Zapier Action". Configure Zapier Let's configure   Zapier   so that it adds a song everytime that the Zapier Action is triggered within Flow. Read more about Zapier Configurations in our   Zapier Docs.   Grow your playlist Done, your bot can now gather songs via channels like Messenger and WhatsApp and add them directly to the Flow playlist.
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
259
In this guide the focus is to show you how to get more people into your Chatbot funnel, we’ll show you 4 different ways to increase engagement. If you want to get the most out of your chatbot it is important to get many people to interact with your bot and to optimize the conversation. Optimizing the conversation is all about personalization, context, and design. Read more in our article about   creating chatbot context. In this guide the focus is to show you how to get more people into your Chatbot funnel, we’ll show you 4 different ways to increase engagement. If you have a valuable bot, the concept is quite simple: More users that interact with your conversational experience means even more value for you and for your customers. To get more people to interact with your bot we will cover the following 4 options: Embed in website Increase button size Add proactive messages Auto-open & Trigger 1. Embed in website Embedding your chatbot on your website can convert your static page into a conversational website and will increase engagement with your bot. Read more in our   Web Widget docs.   2. Increase button size A simple but effective way to increase the number of interactions is by increasing the chatbot button size. This can easily double the number of interactions. Have a look at our   Alternative Button docs 3. Add proactive messages Pro-active messages allow you to share information without opening the chat widget. It is often used to let users know that help is available via the Web Widget. Read more about   Proactive Messages.   4. Auto-open & Trigger With the   Auto-open feature, the Widget opens automatically when it’s loaded. Combine it with   Triggering an Open event   within your Flow Design.      
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
222
One of the most simple but yet powerful features within Flow.ai are conditions. You might know them as if/else statements. With these statements you can create advanced cases in a simple way without coding. If/Else statements    Conditions, or if/else statements, are very simple, no advanced AI and NLP here. Let’s write down an if/else statement: “If you are hungry, then you grab some food, or else you keep designing awesome Flows.” In the same way you can combine different statements: "If you are hungry and thirsty, grab some food and something to drink, else keep designing Flows." You encounter this kind of logic everyday, in human decision making as well as chatbots and can definitely add value to your conversational experience. How it works To create condition in Flow.ai you can drag and drop the condition trigger onto your canvas. To set up the condition, select your condition and add your rules on the right panel. This panel allows you set up your rules and combine rules.   You can really create the great conversational experience by incorporating one or more of the following in your conditions: Parameters Channel Tag User Business hours Let’s have a look at how you can incorporate those in your chatbot design. 1. Conditional content based on answers/parameters Let's kick it off with conditional use case that is very broad. Within the Flow.ai platform you can capture and extract user input and store this in a parameter. You'll end up with a parameter that contains the input of the user. With conditions you adjust your reply based on the parameter that contains user input. Let's have a look at an example where the intent is "order". At the Flow.ai restaurant you can order food and drinks and we set up the following conditions: If the user orders food, then we ask for a drink If the user orders a drink, then we ask for food If the user order food and a drink, then we proceed to the checkout     2. Omnichannel support When you're connecting flow.ai to multiple channels such as Web, Messenger, WhatsApp and Telegram, you will notice that some of the channels support more components than others. Facebook Messenger and the Web Widget supports Cards and Carousels while these are not support within e.g. WhatsApp. With conditions you can tailor the experience to each of the channels.   3. Tags for Marketing Within Flow.ai you can add tags to people which is great for marketing. If a users is interested in women's clothing, simply add a tag. When that user asks for new releases you can directly show relevant content.     4. Profile Information If your bot is connected to messaging platform like WhatsApp or Messenger you will receive profile information that you can use. You can use this information to   Create Context (see 3. Profile Information)   and you can use conditions to alter the workflow. In the example below the user did not receive a confirmation e-mail. If the e-mail address of the user is know we can check his latest order and re-send a confirmation e-mail. Else, we ask for the user's e-mail address.   5. Business Hours Conditions in combination with Business Hours works great when there is takeover required. If it its within business hours you can follow up with livechat. Else, we create a ticket to follow up the next day.    
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
252
Use buttons within your chatbot to set variables or parameters and easily follow up with conditions. In most cases buttons within chatbots are used to send text or to trigger event. You can also use buttons to set certain variables or parameters within your conversational design. This article will show you how to add parameters to buttons or quick replies and how you can easily follow up with   Conditions. How it works   To set parameters by using buttons you can use buttons within a card or quick replies. Select the button and the panel on the right allows you to add: Label: the name of the parameter Value: the actual value that you would like to add for that parameter You can combine these parameters with   Conditions   to easily alter your workflow based on the value of a parameter or use actions for API calls or other integrations.    
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
199
Are you sending out emails to gather feedback and reviews? With an average click-through rate of 2,5% that requires a lot of emails to get a proper understanding of your customers’ thoughts. Can you imagine sending out 10.000 emails and only 250 people will click-through? And from those, only a small percentage will finish your form. If this sounds familiar, then it’s worth implementing conversational reviews. Integrate with survey tools Don’t want to create a survey bot yourself? Want more flexibility, advanced analytics or have a preferred solution? Easily integrate Flow with specialized survey tools and platforms like   InSocial,   SurveyMonkey   and others. What are conversational reviews Conversational reviews allow you to automatically gather feedback in a conversational way by using chatbots on messaging channels such as WhatsApp, Messenger, and many more. These types of reviews are more interactive and allow you to improve the experience by tweaking your design. How to improve conversational reviews To get the most out of your conversational reviews you can tweak your design. By tailoring your chatbot design to your customers you can optimize your experience. Ever thought about including these 3 in your design? 1. Personalize the experience You don’t randomly ask for feedback. There is a prior process, step, or trigger. That could be the purchase of a product, contact with customer service, or attending an event. Use that knowledge to personalize the experience. Something like “Thank you for your purchase and enjoy your new MacBook Pro, I would like to ask you three more question to further improve our service” will be more personal than “Share your thoughts”.     2. Work with Timers Timers can increase responses when using correctly. When your customer doesn’t answer the question after 8 hours the timer will be triggered and you can follow up. It is important to be customer-centric here. Don’t let your bot be that annoying salesperson that keeps on calling. Instead, try to rephrase the question and make it easier for your customer. Instead of an open question you could know ask a closed variant of the question and add quick replies if supported by your messaging channel. 3. Handoff & Handback A chatbot that can hand the conversation off to a human agent in a proper way is very useful. Ever thought about handing the conversation back from agent to the bot? After the agent (or bot) answers a question, that is the moment to ask for feedback. You are already in contact with the customer on a messaging channel. Use that moment and to increase responses. The agent can directly trigger specific events in Flow to start the conversational review. This is support within our own chatapp as well as our   Live chat integrations.  
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
418
How popular is your bot on different channels? How to count sessions per Channel or Widget In this guide we will show how to use the analytics export to count the session per channel. This works great if you would like to get more insights in the popularity or the different Web Widgets, Messenger channels or other integrations that you use. A sessions is considered as a successful interaction between your app and a client, regardless the number of messages.  After creating a conversational experience you can integrate it with different channels. You could add your chatbot to multiple websites, Facebook Messenger pages or WhatsApp. We will now show you how to get insights in the popularity of each of these channels by counting the sessions in three steps: Export your Analytics Download our Template Copy data Export analytics Let's start by gathering data. Login to  the Flow,  and select Analytics from the menu. Click the export button and select a data range. You will receive an e-mail that includes a download button when your download is ready. Download your files and unzip the folder. We will use the "sessions.csv" file. Download Template You can use the previous export to count the number of sessions per channel. To handle the data you can   Download our Template   in Spreadsheets or you can follow up with other software. In this guide we will continue with our template. Download Template The template consists of three separate tabs: Instructions, Data and Dashboard. The Data tab is the only one that we'll be changing. Copy data We can now combine your export with our template. Open "sessions.csv" from your export and select all and copy the data. Open the "Data" tab from the downloaded template. Select all and paste your data Visit the "Dashboard" tab from the downloaded template and you will see the sessions per channelId Which channelId is which channel or Widget? The channelId can be retrieved from the Integration section within the Flow app . Select your integration and copy corresponding link/URL. Find you URL https://app.flow.ai/5b55dcaf -8 f84 -4 f3c-bb6b -1 b022bb78940/channels/ab86e27a -8901 -4 d77 -8536 -bf30f601acc4 channelId ab86e27a -8901 -4 d77 -8536 -bf30f601acc4 Channel Name Test Widget    
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
158
Having several intents that look like each other is dramatic for the performance of your AI-model. Prevent a design where one expression is part of two or more intents. Add training data that doesn't overlap or work with entities. Intents Training data of one intent should not overlap with training data of another intent or as less as possible. The training data should be relevant and specific. Entities Entities are words or small groups of words that are relevant for that particular conversation. They can be names of people, organizations, cities, products, brands etc. These entities are usually   extracted from the sentence   to use later on for an API-call, product recommendation or to redirect to the right flow. Different entities can have the same intent Let's say we are an e commerce company and we sell all kinds of consumer electronics. Instead of creating an intent for each separate product we use entities. A common Intent would be something like "Order Product" and corresponding training data:   Follow up with corresponding Flow Let's say we've got a different process for ordering a Playstation 5 as it not released yet but also a different process for TV's and headphones. Quick Tip: Create entity lists to add synonyms such as television, monitor for the TV category. Use actions and a bit of code to redirect to right Event. You could also create separate Flows for each of the categories. Have look at our   Code Best Practices   for more info. async payload => { // Check if Param exists if ( Array . is Array( payload . params . product_type ) ) { product = payload.params.product_type [ 0 ] .value // Trigger event that matches param name trigger(product) } } Your design will now look something like this:   Now go ahead and give it a try. Your bot should be able to redirect to correct event based on user input.     Intent training data guidelines Use correct spelling Slang/dialect is allowed Be specific Add a variety of words Add a variety of phrase structure Don’t add paragraphs Use the language of your target audience. In most cases that is not too formal and avoid jargon Don’t use intros and outros. Add “I want to order a TV" instead of “Hello, I want to order a TV. Can I do that? Thanks”.  
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
255
When you’re trying to capture a message from a user you may encounter a scenario where they answer in more than 1 message. When your bot proceeds with the next step without letting the user finish their message you might lose useful information. Therefore, it is recommended to create a different set-up. Especially for cases where your are gathering information to push it to a customer service tool. In this guide, we propose three solutions to give the user the chance to finish their message before proceeding; Ask for 1 message Indicate when finished Wait for input Ask for 1 message This solution is very primitive and simple to implement. The downside is that this solution won’t be perceived as a natural conversation. How it works: Within the Bot’s text reply indicate to deliver only 1 message. Indicate when finished When you want to give your customers a bit more space you can allow them to send several messages. This leads to more complex design. Even though the user should indicate when “Done” it feels more natural than the first solution. How it works: The parameters that you can fill are lists. Multiple messages by the user will add messages to the list. When the user types “Done”, we know that the list is completed and we can continue. You can use the following code to check if the user types "Done" and redirect to the right event. Make sure to adjust the names of your parameters and events. async payload => { if ( Array .isArray(payload.params. message )) { message = payload.params. message [payload.params. message . length - 1].value //console .log( message ) if ( message == "Done" || message == "done" ){ trigger( 'Done' ); } else { trigger( 'Question' ) } } }   Wait for input The third and most natural solution gives your users time to create several messages. The design is more complex than the first solution and is also more error prone than the previous two, especially when there is a large difference between the type of users. One person can type much faster than the other. How it works: Similar to the previous step we add several messages to a list. Whenever the bot does not receive a new messages within 5 seconds (or more depending on your case) we can assume that it’s safe to proceed to the next step.  
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
185
When designing a flow with multiple options you might want to branch or split your flow. This article will show how to branch your flows by combining multiple triggers. 1. Create a Flow Create a new flow where you would like to create a branch. In most cases this will contain a question prior to the split or branch. In the example below we will be working with the question: "Would you like to know more about branching your flows?". 2. Add Triggers The closed question above will result in a yes or no input by the end-user. Our next step is to make sure that the user-input will trigger the bot. Therefore, we drag & drop two text replies next to each other.   3. Add content After adding both of the triggers you can now follow up with a reply by the bot in both branches and your done. You've create a branched flow.   Tip 1: Combine branching with Quick Replies     Branching flows and Quick Replies are a great combination. Improve the speed of interaction and reduce errors by using quick replies. Read more about quick replies at our   Quick Replies article Tip 2: Expand and collapse branches           While branching your flows, you can change between views. The expanded view is great for visualizing but takes more space on your canvas. The collapsed view is great for narrowing down your view. You can switch between views by clicking the "> <" sign in your expanded view or by clicking the "<-|" sign in your collapsed view. Tip 3: Multiple branches vs. linking flows Branching is great but branching to often within the same flow might affect the readability of your design. You might want to branch in some cases but link to other flows in other cases.
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
197
Capture user input to create tickets, make API-calls or to integrate with external services. Captured data can be used to: Create a ticket in your CRM tool Make API-calls Integrate with external services Entities vs. Any Text The two most common ways to capture user input are by using Entities and by using Any Text. Entities Entities are variables you give your bot to categorize and interpret user input. Entities are great for picking variables out of sentences such as arrival city or destination city. Any Text The Any Text Trigger can be used to capture any text that the user types. This option does not require any AI and therefore doesn't need training examples. You can combine Any Text with Actions to   Check user input. Use extracted data or input Your parameter has now been created and contains a value. You can use this data to send an e-mail or to make API calls for several solutions. Have a look at our   Code Actions Best Practices. Send e-mail with extracted data To send an e-mail with the extracted data you can use an action. In the example below the following three parameters are incorporated: name email message Make sure to extract them before sending the e-mail. Otherwise you'll be missing some valuable content. async payload => { var name = "-" var email = "-" var message = "-" // Check if the param "name" exists if (Array.isArray(payload.params.name) ) { var name = payload .params .name [ 0 ] .value } if (Array.isArray(payload.params.email) ) { var email = payload .params .email [ 0 ] .value } if (Array.isArray(payload.params.message) ) { var message = payload .params .message [ 0 ] .value } toolbelt.email({ to: 'yourname@yourdomain.com' , // <-- Vul hier een email adres in subject: 'New message via chatbot' , message: `name: ${name}, <br/> email: ${email}, <br/> message: ${message}` }) }  
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
252
Are you looking for a way to change the language of your bot? You can change the Brain of your chatbot from within  the Flow App . Select Project settings Access Brains Select your language   Supported languages At the moment the following languages are supported: Abkhaz Afar Afrikaans Albanian (ʃc͡çip) Amharic (ኣማርኛ) Arabic (العربية) Armenian (Հայերէն) Assamese (অসমীয়া) Aymara (Aymar aru) Azerbaijani (Azərbaycan dili) Bashkir (Башҡорт теле) Basque (Euskara) Bengali Bhutani Bihari Bislama Breton (Brezhoneg) Bulgarian (Български) Burmese (မြန်မာစာ) Byelorussian Cambodian (ភាសាខ្មែរ) Catalan (Català) Chinese (汉语/漢語) Corsican (Corsu) Croatian (Hravtski) Czech (čeština) Danish (Dansk) Dutch (Nederlands) English Esperanto Estonian (Eesti keel) Faeroese Fiji (Hindustani) Finnish (suomi) French (français) Frisian (Fries) Gaelic (Scottish Gaelic) Galician (Galego) Georgian (ქართული ენა) German (Deutsch) Greek (ελληνικά) Greenlandic (Kalaallisut) Guarani (Avañe'ẽ) Gujarati (ગુજરાતી) Hausa (هَرْشَن هَوْسَ) Hebrew (עברית) Hindi (हिन्दी) Hungarian (Magyar) Icelandic (Íslenska) Indonesian (Bahasa Indonesia) Interlingua Interlingue Inupiak Irish (Gaeilge) Italian (Italiano) Japanese Javanese (baṣa Jawa) Kannada (ಕನ್ನಡ) Kashmiri (कॉशुर, كٲشُر) Kazakh (Қазақ тілі) Kinyarwanda (Ikinyarwanda) Kirghiz Kirundi (íkiRǔndi) Korean (한국어) Kurdish (Kurdî) Laothian Latin (Lingua Latina) Latvian (latviešu valoda) Lingala Lithuanian (lietuvių kalba) Macedonian (македонски) Malagasy (Fiteny Malagasy) Malay (Bahasa Melay) Malayalam (മലയാളം) Maltese (Malti) Maori (Te Reo Māori) Marathi (मराठी) Moldavian Mongolian (монгол) Nauru (Ekaiairũ Naoero) Nepali (नेपाली) Norwegian (norsk) Occitan Oriya Oromo, Afaan Pashto (پښتو) Persian (Farsi) Polish (Polski) Portuguese (Português) Punjabi (ਪੰਜਾਬੀ, پنجابی) Quechua (Kechua, Runa Simi) Rhaeto-Romance Romanian (limba română) Russian (Русский язык) Samoan (Gagana fa‘a Sāmoa) Sangro Sanskrit (संस्कृतम्) Serbian (српски) Serbo-Croatian Sesotho Setswana (Tswana) Shona (chiShona) Sindhi (سنڌي) Singhalese Siswati (Swati, Swazi) Slovak (slovenčina) Slovenian (slovenščina) Somali (af Soomaali) Spanish (Español) Sudanese Swahili (kiSwahili) Swedish (Svenska) Tagalog Tajik (тоҷикӣ) Tamil (தமிழ்) Tatar (tatarça) Tegulu Thai (ภาษาไทย) Tibetan (བོད་སྐད) Tigrinya (ትግርኛ) Tongan (Faka-Tonga) Tsonga (Xitsonga) Turkish (Türkçe) Turkmen (Türkmen dili) Twi (Akan Kasa) Ukrainian (українська мова) Urdu (اُردُو) Uzbek (Ўзбек тили) Vietnamese (Tiếng Việt) Volapük Welsh (Cymraeg) Wolof Xhosa (isiXhosa) Yiddish Yoruba (Èdè Yorùbá) Zulu (isiZulu) Level of accuracy for the NLP to match You can change the level of accuracy for the NLP to match. When shifting the threshold to a higher score, the AI-engine will match user input to a flow with more confidence. Increasing the threshold might result in more "Unknowns". Minutes to auto-pause Within your design, you can use actions to pause the bot. The variable Minutes to auto-pause indicates the of minutes a bot is silent when it is automatically paused.
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
198
We've built the Web Widget in a way that it keeps its context. If your customers come back for a visit on your website they can easily continue the conversation with your bot. That is great for the customer experience but when testing you might want clear the widget. You can, of course, use the Try it out but that design is a bit different from your actual Web Widget design. Let's look into different options that you have to clear the Web Widget. The first 3 can help you out for testing purposes. The fourth one shows you how to change the standard behavior of the bot. 1. Incognito Tab The first one is very simple. Copy the URL of your Web Widget and past it in an incognito tab in your browser. The Widget will not recognize that it's you again and will start the conversation from scratch. 2. Clear button A second option is to use the reset button within the Web Widget. You can enable and disable this within the Customize section of the corresponding Web Widget.   3. Reset Trigger A third option is to build a reset within your Flow design. Your chat conversation will not be deleted but it resets all the parameters so that you can start all over.   4. Session Storage By default the widget stores messages across browser sessions. By setting flowai_webclient_storage to session you can change this behavior. When set to session all messages are removed when a user closes the web browser. See our   Web Widget docs   for reference.
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
182
Chatbots are being used in almost every type of organization mostly in customer service, marketing, and sales or as a virtual assistant. In this guide, we will show you how to create such an AI chatbot in Arabic with right-to-left text alignment. After creating your project you integrate your bot to different channels such as Web, Messenger, or WhatsApp. Set language to Arabic The first step is to set the language of your chatbot to Arabic. You can pick your language when creating a new project or you can pick a language from within languages in Project Settings. After setting your language to Arabic, your design will automatically handle right-to-left text alignment.     Design your experience Now, we can start building the AI chatbot. To do that you can create a new flow and start with a trigger. In this example, we’ll stick to basics and we start with a text trigger followed by a text reply.     Test your experience After designing your experience you can test and integrate it with different channels such as Web, Messenger, WhatsApp, and many more. You can use the built-in Try it out or you can add more channels from within the Integration section. In the example below you see the Try it out and the web widget.  
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
171
There are several reasons for creating more than 1 chatbot. When managing different bots or projects and sometimes even for different companies this can be challenging. Three common areas where you encounter more than 1 project for 1 or more organizations: 1. Working as a digital agency As an agency, you work for different brands. For you, as an agency, it should be easy to switch between these different environments. Within Flow.ai our   agencies & partners    are important. So, you can be part of different organizaitons. By using the drop-down menu on the top right you can easily switch between your brands. 2. Different use cases Within 1 organization there could be different use cases to incorporate a chatbot. On the support page there might be a customer service chatbot, while you might use a different bot as a lead generator for marketing & sales. In Flow you can easily change projects from dropdown menu on the top-left. 3. Testing purposes After launching your first bot, you might want to try out new features and experiences. You don't want to take the risk of modifying your launched bot so it might be good to create a new project for testing. You could even import your original bot by using the   importing your project. Happy Flow-designing
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
170
Within you design you can use buttons to enrich the user experience. Buttons can be used as a single button, within cards or within carousels. The example below shows you how to add buttons to a card. 1. Add Button To add buttons to your card, select the card on your canvas. Select add button from the panel on the right.     2. Customize Button Select the button from your canvas. The panel on the right will now change and you can add the following: Label - The name that will appear on your button Type - What would you like the button to do? Value - This is the value that corresponds with your type. This can be a URL, text, event & more     3. Re-order Button To re-order your buttons you can select the button that you would like to re-order. Then drag & drop the button to its new position.    
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
191
Save time by copying or duplicating your existing flows. You can duplicate entire flows and edit the copy to save time and design your conversational UI even faster. Select the flow that you want to copy or duplicate. Then select "Duplicate" from the drop down menu within your flow.  
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
198
You can use the Import and Export functions from  Flow   to copy or duplicate your existing project or chatbot. Duplicating your project can be broken down in to three steps. 1. Export existing project To export you project, select Project settings, Backup and click on export. A download with your project will start. 2. Create a new Project Now create a new project by selecting Add Project from the drop-down menu in the top-left corner. 3. Upload project Select Project settings within your new project. Continue with Backup and Import. You can now upload you downloaded file from step 1.    
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
189
People are not perfect, the same goes for chatbots. Even though you spent a decent amount of time on designing, testing and you are using advanced Deep Learning techniques, your bot will fail. Make sure to prepare your bot with a decent fallback and set up an environment so that you can learn from these failures. In this guide we will show you: Best Practices: handling the unknown 3 Ways to handle the Unknown Best Practices: handling the unknown       What do you, as a person, do when you are prompted with a question that you don’t know the answer to? Well, those steps are quite similar to the process of chatbot. 1. Acknowledge failure As your parents told you not lie, you will be honest and acknowledge the “failure”. Same goes for your bot. When your bot has no clear answer to question you can start with acknowledging the failure with something like “Sorry, I didn’t quite get that”. 2. Propose a solution or workaround Now, we’ve acknowledged the failure but there is no solution at all. So our bot is definitely not helpful. Always try to propose a solution. 3 Common solutions that will be discussed in the next section are: Asking to rephrase A handoff to an agent Extract topics 3. Gather and process feedback Now, we’ve proposed a solution but still, how do we know if that answer is satisfying? Simply ask for it. A question like “Was I able to help”, followed by a thumbs up or thumbs down button can do the job. This is a great opportunity to learn how you can improve your bot. Read more about it in our article about   The Feedback Flow. 3 Ways to handle the Unknown Allright, that’s that part. Now we can start building an actual solution for our bot to handle the Unknown. You can use one or a combination of the following techniques to handle situations that are not known to your chatbot. It all depends on your case and available resources and agents. 1. Rephrase and show topics A simple solution is to ask the user to rephrase their question. Even though this solution seems simple, it can be very effective. In addition to rephrasing you can also show topics within   Quick Replies.     2. Handover A second solution is to hand the conversation over to a human agent. Agents are very good at understanding complicated questions and answering those out of the box. Acknowledging failure is optional in this scenario. It's does not add any value and there will be a person following up anyway. Two common handoff scenarios to consider are: Live chat Ticketing software Following up with livechat or creating tickets in a CRM all depends on your existing process and available resources and tools that you are using like Hubspot, Salesforce or Freshdesk. Zapier is a great tool that you can use to connect your bot with many other solutions. Read more about   how you can improve your chatbot by using Zapier. 3. Extract topics A third solution is trying to extract topics that the bot is trained for. When the bot is not able to recognize a topic we can still hand the conversation off to a human colleague. To set this one op we'll first need to transfer the input of the user into a parameter. You can do that with an Action with the following code that will create a parameter with the name "query". The value of this param will be equal to the input of the user. async payload => { return { params : { ...payload.params, query : [{ value : payload.query }] } } } With our new param in place we can follow up with   Conditions   to see if there are topics or words that we recognize. If there are topics that we recognize we can show the user how the bot is able to help him with that topic.      
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
284