 Most of you would have been here yesterday and I'm a little nervous because Nareesh Kickstarter the Kickstarter the day with saying that no no collaboration and I'm here to talk about the power of collaboration so but I have to repeat what Nareesh said look at yeah so that's what I'm talking about but that's the easy bit that you get out because what gets stuck that gets in your brain so take it with pinch of salt take it with the context this is what work for me for for quite some time for my company and if you are in the similar context take it and try it if it doesn't work don't blame the tool move on and try something else okay so I'm chatting with that and Kalpesh had a very energetic talk so I'm going to some of the talks there is a lot of there is a lot of overlap between what he spoke and what I'm going to speak so there can be differences also but again take it with with your contacts try it with your contacts and see what works experiment it and and then see okay what is working what is not working don't blame the tool never okay so I'm going to share with that so this talk is about impact mapping which is a technique which is mentioned in the book with the same title called impact mapping and this is about how do you deliver software which creates some impacts when we are talking about impacts impacts are not just some numbers impacts are we are talking about actual behavior change of the users of the customers so and which is again based on the on the philosophy that of which demo clouds that is when you when you actually put a lot of people together and when they know that why they are working what they're working then they usually good ideas come out and it works so why do we need something like that so let's look at some some data why why do we need something like that and this is a standard case report I'm sorry that it's not very clear so which which they shall be company this they publish the success of software projects or rather the rather the outcome of software projects and you can see for the last five years only 30% of the project has been successful and others are either in the challenge stage or pain state so when you say challenge what does it mean it means that either it was overrun by budget horribly or or over or it didn't meet it was not in any alignment with the market when that is what it's a challenge statement and this talk is rest of the talk is about how do you avoid that kind of thing how do you create alignment with the market and then how do you do that at the company level that's the rest of the talk is about so this is one path this is one data and how we are familiar with this particular model have you heard of what has come for calpish folks walk about that the the the previous talk so this is a bunch of ideas come from either a PMO or or or a or a you can call them as the highest paid people in the group and they come up with a bunch of ideas they get ideas and then they come and throw it to the throw to us a set of group of people which is called the so-called agility they work iteratively their their job is to pick items in the back up that's it and then there are there are demos at the end of it but after the day after the demo it doesn't go to the production after the prolonged period of time it goes to the production and that is when it's fallen to the users and that is when we realize that okay nothing is working and this is this model is called the waters come for model where the problem is that the iteration is only within the middle of the group the iteration doesn't happen with the entirety and and this is this is one problem where the alignment lack of alignment becomes a huge problem and to avoid that people create a lot of documents and check into the repository the documents such as the vision document the project chatters so the problem with that is that it is a shared document it's just rest in the in the in the repository I hardly get updated with new learnings and it doesn't result in a shared understanding okay and if you are still not convinced with what I'm talking about this is a test for you ask your team why are we doing what we are doing then you yourself will be surprised by the answer that you give and you will be surprised with the answer that you get from your team members and this is this is a this is a common problem this is not unique to a specific theme or something and this is a unique problem and this is why we need certain tools to help us or take certain techniques to help us bring in shared understanding among the team and so how do you do that that is where impact mapping comes into picture it's a book written by Goyko I'd sick with the same title and in that what what he talks about is about how do you bring in alignment within the team but how does it do that it it uses the wisdom of crowds you collaboratively do that you create create creative ideas from from all types of people rather than just a group of people and then the another beauty of the system is that it relies on questions and this is where lead has grown again and again if you ask the right question there is high probability that you will get the right answer so the entire session is about asking the right question mainly the the four questions that it forces you to answer four questions the four questions are first is why why are we doing it next is to whom we are doing it and the third one is how are we going to do that and the lastly what are we going to do okay and the last what when we answer what that is the time when the feature backlog comes up until then feature backlog doesn't come up and that is the that is the lowest priority item in the list what has to change over a period of time okay so and then the next question is who should answer this ideally everyone in the team because the idea is to have shared understanding to know that everyone in the team knows why they are doing what they are doing and but if the team size is too big then you need you can you can decide who can be there but the ask that you can use to decide is that there should be decision makers in the room decision makers who usually architects can make technical decisions so they have to be there in the team and business people who can make decisions on behalf of the business if you don't have decision makers within the team then that means this is not going to create any kind of outcome because we have to make decisions during the week okay so this is a two overview about about impact mapping so I'm going to use our stories our experiences with while doing business I work for a company for my journal we are a consulting company so we had we have been trying this with multiple customers now and we have seen the results that the reason I am here to share my experience so this is an experience report about impact mapping that we had and for this particular experience report let's go to Gujarat Gujarat is a state in in India those who don't know and there's a district called Sabarkhanda so what happened was the district development the district development officer of Sabarkhanda approached us he's an IAS officer he approached us to build a MDM solution which is a mobile device management system for for their health workers so mobile device management system is remotely controlling your mobile devices from a central location so this is very common in fields in businesses where they have field staff for example amazon's flipkarts they have a lot of field staff and they will have some kind of a remote control mechanism so that they can control these devices remotely so what happened was they have health workers in the in the health department they go and visit the villagers to for to their purpose is to one is they have to collect the health information about the villagers and so that and then they collect that data and bring back and sync it with the cloud and this this this will be used by the medical officers to analyze the problems in the village and create and suggest better medicines or precautionary measures and another one is to create awareness among the villagers about malnutrition child mortality and vaccination etc so what they have is in these devices that this is a actual picture so in the devices that they carry the Android tablets they it will have videos and powerful presentation which they show to this particular video these ages to create awareness about these these issues they don't take proper vaccination they don't they are not aware of a balanced diet and through this video they create those awareness and what happened was when the devices reached around 300 400 devices when they when it reached that scale they started losing the the track of the devices and not only that they wanted to actually control these devices for example the content in this device has to be updated when new content new new video informational video come into picture comes in especially units of releases it very frequently they want to make sure that these videos are they are updated in the devices in as early as possible so if you are doing it in very manually then it takes a lot of time so that is the reason they approached us and then we realized okay rather than MDM as a mobile device management system has a set of redefined features rather than just jumping on to that let's see okay what problem you should first solve because rather than just building features we feature and then if it is not solving all the problems that they have in the field then there is no meaning in doing that so just stepping back and coming back to impact mapping impact mapping the beauty of the impact mapping is the center of the map so it's a mind map we will create it as the top goes so center of the map is why we are doing why we are doing what we are doing and it requires a lot of respect to come up with that you can't rush it so what is suggested is that you split the session you split the creation of the map into two different sessions firstly understanding the why and then filling the rest of the map so the first session is called prepare the map in both the session use the yardstick to decide who should be there in the team especially the decision makers has to be there in the team and then who can actually who can help us explain the process is what we need in the team ideally everyone in the team if the team is too big then you think of how do how do you make sure that the decision makers are there in the meeting so the first session is about preparation of the map and as a first step in the preparation of the map you have to understand the goals so how do you arrive at the goals you can just throw ideas that that can be done or another way to do is coming with the using the shopping list shopping list is a term goco uses in the list shopping list is nothing but your features so from that you arrive at the goals but how do you arrive at the goals using five eyes so the beauty of the system like I mentioned earlier is it's based on questions you are the right question there's a high probability probability that you will get the right answer so you you do five eyes and then you arrive at okay why why why and in the end you know okay what is the reasoning behind this particular teacher sometimes you know that okay there is no reason to build this and that is that should not be there in the backlog at all so again based on conversation so this is a conversation that we have okay why do we need why do we need a registry of monitors or registry of devices we don't know which devices are being used by phone okay so similar conversations happen and in the end we could come up with goals so we arrived at four goals or if you look at very closely they were facing two kinds of problems one is that they don't know what is happening in the field they need they don't have any kind of monitoring mechanism saying that are the devices being used properly are the apps that is which is installed on the devices are those being used properly and the next stage actually updating the devices when the new version of the app comes out it has to be updated as early as possible when the new content comes out that needs to be updated it has to be updated as early as possible all all all remotely so you can you can see two types of problems one is actual remote control one is actually monitoring the system and then the so no no you know these are the problems but how do you know that the solution that we are going to come up will indeed solve this problem that is when where the measurement comes into picture so you know you need to have enough measurement to track that okay is it really working or it's not working so to to create good measurements there are five aspects one is a scale that is what are you going to measure next is a meter how are you going to measure it and the benchmark that's a current situation and what are the investment that you're going to make the it can be the course or the time and lastly the target what do you expect at the end of the experiment and our idea is to reach this particular target so the measurement the goal will work once it reaches the target which is not then there is a failure so we converted that measurement the goals into a table of format answering almost all of the all of the measurements yes it doesn't have all the all the things I mentioned in the previous slide it doesn't have constraint and the and the and the benchmark benchmark is zero because they don't know anything that that's the current situation and constraint we haven't put that because at that point of time we didn't have enough data to define that so it's okay if you don't have it at that point of time it's okay you should have it at the latest point of time as as as you move forward you should have it so ideally you should have all the numbers filled in but in case if you don't have don't don't wait until you have all the numbers you move on and so now what is the problem you have four problems to be solved and you know how we are going to measure that but the the reality is that we can't solve all the problems at once you have to choose one the choose what you what you're going to solve first and that is the next step of impact mapping that is planning your first milestone so usually when you have conversations within the team you within the right people the experts or the business people you know okay what is the first problem to be solved if you don't have that you're still finding it hard to delight it you can use these techniques either a voting mechanism or a virtual cash virtual cash I think you might be already familiar you give virtual cash to people and see okay this is a value of this particular thing and which has the highest value that gets selected and this is also a collaborative thing there is no nothing like okay one person's voice gets gets up voted I mean this this is across the team so so usually you will arrive at the right ones at the end of this exercise but you what in my experience with the conversation you you should have a clear winner at this stage if you're not if you're finding it hard then use these techniques to arrive at the right milestone so this is where the preparation ends when you have the first milestone identified you have the the center of the map created so we did a little differently over here so in the previous slide you can see that we had four problems and so at the end of this exercise we realized that okay what we need first is to identify the monitoring thing okay how are these devices being used in the in the field once you know that we can put appropriate remote control mechanisms so that's how we approached it and this this is we are still working with the customer so we'll still in the collection data collection stage we are still deploying the app to the more and more devices so what we did differently was after this we created a single goal that is first we have to identify the usage pattern of hundred devices so we are not targeting all the devices hundred should be enough for us to know okay what is happening on the field so we combine both the things together those are measurements which we help us to know that okay are the devices being used appropriately or it is not if it is not being used appropriately what are the reasons it can be the network or it can be the lack of knowledge of the health workers because they are not savvy people so it can be a training that might be required next so again we don't need software to solve that but software will help the product will help you know okay what is happening in the field rather than all the medical officers going and checking what is happening so that is how this particular process is still undergoing we are still learning from the field we have found new learnings as of yesterday day before yesterday we have got new surprises from the field and that's what is expected okay so once you have the first milestone that becomes the center of the map and this answers why are we doing what we are doing okay so just a little more about why this is important so Gary Flynn who was a research person in the field of naturalistic decision makers so he did a lot of studies among critical who works in the critical field such as such as critical care measures or say firefighters or the operators at the nuclear power plant or chess masters so what he started what he what he wanted to find out was how do they make decisions on a split of a second because they don't have much time and what what he found was okay if they if you know if humans know why they are doing it they can make decisions first usually we get confused because we don't know why why why why we are doing it so yes as software professionals we will not work on high critical systems that's true but what is common in software is unexpected events even if you do the same thing and again and again at times you get you get surprised with what is happening so how do you make decision when such unexpected events happen if you know have a clear goal of what you want you usually make better decisions so that that's why a lot of emphasis on coming up with the center of the map so and it requires a separate session and it is better to be done in that okay and the idea is to have a shared understanding so so once you create that now you remember the four questions the first question we answered why are we doing it we have to answer the rest of the four that is the next is the first level of first level of map is cool and and again emphasizing that getting the why correctly identify the value that you want to deliver and usually the value value it's not about the features don't look at the features or the scope look at the value that you want to deliver and next is who who are who are the actors in the system we are very clear about the end users but they are not just the actors in the system there are so many people involved in your software and so again rely on the right questions to answer that the as the questions to be asked are who gets impacted with this that's the most obvious question who can disrupt this that's also very important and so that we can avoid that that disruption and who can help us reach to the desired desired effect so to give give an example this is what we came up with for our MDM product the villagers are the obvious impactor they get impacted with with the with the product who can help are the health workers and the medical officers by using the devices properly and medical officers training the health workers to use the devices correctly and network if this completely relies on mobile network if that doesn't work then there is a huge problem so they can disrupt from achieving what we want and there are apps that is being used by the health workers the user experience of the app is not right then that also gets affected and keep in mind that these health workers are not at all tech savvy that's why it's more important over there any any product user experience is a big thing but especially in this kind of situation it has to be very very easy so so these are the people who gets who can help us or who can disrupt from achieving what we want and how do you get this right ideally we should have personas how many of us are familiar with personas okay so i didn't include that this in this presentation because that itself is a separate talk by itself so ideally i should have shown that shown the persona instead of the of the role but but it might get confused that's the reason i didn't use that so personas are the best way for you to empathize that with that particular role so when you say personas you need to know exactly what they are doing what their day looks like okay what what interests them what frustrates them you need to know exactly everything about that particular person so if you know that you will be able to visualize that person better and then when you are developing things you can see how that person will be using so it's it's it's a very good exercise highly recommend that when you're doing when you're doing back mapping come up with personas and then stick that into the informative workspace so that people can relate to that and it works very very quickly so that's a that's something that i've not included over here but i highly recommend you to do that and so we have the first level so we know why we are doing it and we know who for whom we are doing it the next level is okay how are we gonna do it and again a set of questions um so the in the how part what is the key over here is the change the change means change in the behavior that is what is the impact means we are not able to create any change in the end users then that means the software is not getting impacted right so um we came up with a set of how okay so the better health for the villagers that is what our aim is okay and all others are related to that but how do you get it right to get it right focus on the activities that person will be doing and what will happen when this particular goal is met what is the change and that change is what is what should be the focus for so the key over here is the impact and that impact has to be a change in the behavior and that is something that you have to keep in mind just so we have identified the how i'm not going to the detail of mdm because that's that's not the point over here point is to actually get across okay how can we use in that nothing and now before this step we can do one more thing you you might be remembering the shopping list every team will have a shopping list of features and you can take that and put under the how i mapped it to the how but that you can do it you cannot you may or may not do it that's not very important but that's one way to visualize okay what are we going to do another thing over here is that you can this is where the where the collaboration comes into picture use the diverge and converge of the design thinking so what what this does is give 20 minutes to the team at time works time to the everyone in the team in in the room and then ask them to write as many ideas as possible one one thing to keep in mind is that there is no no criticism about any of the idea every idea is good and let them come up with as many ideas as possible and there is no talking at that time point of time this is where the creativity comes into picture and this is a very powerful thing in in design space so this design studio that kalpesh was talking about in the earlier talk or the design thinking actually encourages to do that okay so diverge and converge so what you have to do is get as many ideas as possible and get the best out of that so time works into 20 minutes and make sure that you are not criticizing or appreciating any of the ideas at that point of time okay the ideas to get as many as possible and how during that process ask this question so that people can focus on the impacts and the and the goal okay so these questions will help them to be very focused rather than coming up with the random idea okay but it doesn't matter whatever the idea is there is no criticism for that and we came up with a set of uh set of features uh we didn't actually look back and said check whether it is actually but part of mdm or not because our idea is to learn the usage so uh we didn't much care about okay whether it is part of the mdm solution or at all or not and now the moment you have a lot of number what is next what is needed to be done next is to actually prioritize and that's the next step uh you need to you can't you can't do everything the idea is not actually to finish all the feature list the idea is to see the impact as early as possible so how do you do that so you have to identify you have to identify the priorities again a set of questions here the questions are uh if you look at the questions uh the ones that is going to give us immediate results that is what you have to uh you have to focus first so the questions are like what are the key assumptions we catch and are there any high value but low hanging low hanging of course in this okay so uh here again you can use the same voting mechanism one more idea over here is to use a color coding uh to to identify those low hanging ones so that you can do that quickly and see whether there is any change you don't have to wait until the end end of uh or for a prolonged period of time to see whether it is actually working so so we did that uh and the the stars represents the votes and we had a set of voted ones so what we did was we focused more on the android app itself less on the reporting because what we realized was if you have uh if you have the data in our database at any course we can get the data that's not a big problem so you don't have to wait until the reporting features are done to go for us to go into the market we can go into the market you will do before that or we can start deploying even before that so that is what we did with this so we hardly have any any features for uh for we initially thought okay we will not focus on the reporting features at all and because the key over here is on the learning so 50 moments and so this is the feature list but keep in mind that as you progress this can change or i would say this should change okay don't keep this as a constant keep the rest of them as a constant and change this according to what is happening when you deploy it so uh so this is by this the map is done uh uh this is a complete mind map the beauty of this mind map is that it covers it conveys you very beautifully this is a picture it's like an image you can stick into anywhere and then people should get what we are talking about or what why we are doing what we are doing and who are the people involved in that and but before that even even after that you can actually do some more prioritization and that is a next major thing that is uh which is called the earn or learn session so in this uh you can ask this question do you really need a software to test this can you do sometimes to prototype to get initial feedback because you know that once you write the code you have to made in it which is becoming very very costly and all of us know the cost of maintaining code right and uh so we don't need as at times if it is if it is a so the question to ask is are we going to learn anything if we are going to learn anything do we really need a software can we do with with something else usually a prototype if you're going to earn something earning is not always revenue it can be cost cutting reduction in the cost then how are we going to do that is there any half half half partially manual process we don't have to automate everything through a partially manual process can we start earning things so these questions are very important very important and the design thinking ideas like design studio or designs plane can help you to quickly validate things so uh and uh with this with this exercise we came to two things we didn't we didn't come up with any prototype because we we realized that we be hard for us to actually test it we want to install it on the device and test it so there is no other way to test it other than deploying the software but one decision we could make is that we don't have to support all the android devices because all of them are using a specific device and most of them are running on a very older version of android so the the diversity of segmentation of android we don't have to worry about over here at least in the initial stages for the hundred devices at least because the idea is to learn faster and whatever we need to do to learn faster that's more than sufficient this became very useful because uh certain apis to to track the app usage the those apis have changed uh the latest version of android so rather than adding code for supporting the latest version we realized okay but let's not support this particular feature the latest latest version of android will support only the 4.4 version of android so those kind of decisions can come when you ask the right question and uh you also decided to lower the priority for reporting that is something that i already know and um okay so um so we're done with the map uh this is a final map and uh again i i i i mentioned this is this is a very simple way to communicate okay what what we are doing and and uh what is expected out of this particular experiment and now the question is okay how how do you measure it so we came up with a measurement initially but that is not reflected on the map yet so how do you do that so there are four uh couple of ways to do that you can add extra nodes within the within the map to do that or you can make it as bullets under below the map but the some of the software that we online software doesn't support bullets or say doesn't support adding nodes so one thing that you can use is each each place you can retrace it with with the measurement that you want to do for example uh in the map in the map that i've shown the center of the map says learn device usage for the hunter devices so that itself shows okay what is the measurement once it reaches 100 okay we have reached an experiment stage and then we can look back and say okay what is happening so you can do that by by expanding your node to reflect the measurement another way to do that is a separate metrics table this also you can draw it in the same uh whiteboard or in a in a software that you are using uh to this will this will actually uh uh actually uh share the information on what is happening on the field to everyone in the team and there is very rare chance that people doesn't uh the team the team members even know this so they know exactly what is happening uh within the field by by by creating that particular thing on the whiteboard itself so this is what we that we did and we we use this same format to share share the report with the stakeholders okay what is happening in the field this is what we shared so like i mentioned we are still under the deployment we have reached uh around 40 to 50 devices we still the device usage is still low we are trying to see whether it's a network problem we are in talks with the network uh people over there to see okay what can we improve and we are also brainstorming okay as it is low now itself should we wait until 100 devices are raised or can we do something else should we stop this experiment and should we move to something else so uh that is what is happening uh and uh so the idea over here is that even before we build all the features we started asking okay we should be really need to build all the features because there is no meaning in building it because it's not creating much impact so uh and just to recap uh this is the mind map of the impact map i don't know whether it be very clear uh sorry about that so in back map uh two sessions of impact map first one is the preparation for the map to identify the why the goals of the of the current current experiment and the second one is actual creation of the map in the preparation step identify the goals identify the measurements and plan your first milestone and in your in the while creating the map um draw the skeleton that is answer all the rest of the question whom are we doing it for why are we doing it for and how are we doing and what we are going to do and ask the question is there any alternate that we can do so that we can get to feel that identify the key priorities and focus on that and lastly do we really need a software to do that or can we do uh can we learn this much quicker and what it tries to avoid is just the typical planning smells where the the the teachers are either prioritized by someone which can be like a feature for that particular person or a for a group of people or some ideas okay just random ideas coming and then that gets added to the backlog and we don't know why we need that and we we stick into the wrong assumption we thought okay this is why we are doing it and we don't we don't validate over the process of building it and in the end we realize okay we have you have much much away from the uh from the from what we are supposed to do and in the end it it it tries to avoid uh creating wrong solution or let me put in this way it will try to fail you early you learn as early as possible are we on the right track or not and that's also a good thing failing is per se is not a bad thing but if you if you are learning from that then it's a good thing um again some of the key learnings uh to emphasize impact is all about behavioral change it's not just some numbers it has to really change someone's behavior especially the users of the system the customers of the system and you get usually get right answers when you ask the right questions but that's a that's that requires practicing the right but do practice it and you will eventually get the right question right answers and always measure uh so that to make sure that if you need a change you can make that change as early as possible make it visible so that uh to improve the shared shared understanding and shared understanding is not about shared documentation it's about collaboration and uh two or three things uh what I learned from this uh is that this one thing that I've I've started personally started feeling is that this is the best way to inject brain thinking uh the the thinking of experiments how do you bring in that kind of a mindset by asking these questions uh usually the team changes their thinking process towards experimentation and they open their minds to try that out which I've been struggling for a prolonged period of time both within the team as well as the customers but this has been successful so far and it has worked for me uh again uh tried out like in your context and see whether it works but usually asking the right questions usually works and um two things that you can use as part of impact mapping is one is design sprint design sprint is uh uh how many of you are aware of design sprint okay design sprint is from google vengers it's a very structured approach of five-day uh sprint where you start with an idea then you brainstorm and come up with a solution which you think may work and you even test it with a prototype and then at the end of the day the fifth day you validate your learning you don't need a working software for doing that and in through impact mapping you see opportunities where you can try it and then you can try this is not just for a specific software complete solution you can try uh design sprint even for a future to come up with better ideas on how to implement it and again diverge converge 20 minutes uh across the team and in the end you work for things and then take the pick one pick the ones which which you think will will uh will empathize with the users more and user story mapping just patterns user story mapping that's the best way you can slice things very very small so when you are arriving at features use user story mapping this also forces you to empathize with the user because the center of the user story mapping is a user it's not the solution it's not the feature it's about the user and that is one thing that i've started realizing user story mapping is something that started very recently but design sprint has worked very well with back mapping and that's that's my experience and that's it and uh i think uh these are some references and i hope you take the learnings from here and start doing in back mapping questions any questions yes yes yes yes yes you can slice it really really well using the user mapping because what what you need is to try out something very very quickly for that you don't need the moment you will start looking at the feature level it is it will grow but when you look at users behaviors or activities then you can get the smaller size slice again it was not for how it was for what yeah that's why the questions comes into picture people are asking the same question so that people are focused towards the activities of the users rather than some just some random idea and anyway you're going to do a converge after diverge so you will get you will prioritize things which are which are mapping to the because it's the center of the map is the goal so it's very rare that you take something which is not mapped very very rare the diverge converge is better than in 20 minutes so that as many ideas come in within that 45 minutes to one hour ours is a very small team so we didn't have to choose team so i think that's it thank you i'll be here if you have more questions