SCIP AMA - Ask Me Anything
IMPROVING THE IMPACT OF CI BY FOCUSING ON THE DISCOVERY PHASE
SCIP: What are the most promising uses of AI in the discovery phase?
GRAEME: The secret of excellent Competitive Intelligence is asking great questions. To do this, you need to get to know the people who will use the end product. You need to understand why they are asking the question. You need to ask great questions and get your decision-maker to give you the questions they really want answered.
Machine learning is beginning to do wonders within the secondary collection part of CI, but this is mainly due to the desire to find the right information quickly.
I am no technophobe, and there are some great tools around at the moment - like Klue, Intelligence2day, Crayon, and others. However, AI is not a super drug. CI analysts still need to think, collect, assess and analyze. We are a long way from AI doing a decent SWOT analysis. Remember that one of the best tools is getting out in the field, getting your boots on the ground by finding out what's happening at conferences and tradeshows, having coffee with experts, and using these opportunities to see what is behind the front gate of your competitor.
GRAEME DIXON - OCTOPUS INTELLIGENCE
Graeme Dixon is CEO of Octopus Intelligence, where his clients include PKF, Samsung, and Tata. Graeme has 10 years of Military Intelligence experience, 20 years Competitive Intelligence & Research experience and is a qualified Professional Investigator. He currently sits on the SCIP Member Advisory Board, is former Chair of the Association of MBAs, and current Chair of SCIP UK.
SCIP: Can you suggest a framework or method to measure CI outcomes from the discovery and requirements gathering phase?
GRAEME: Once you have answered the questions, done the analysis and reported the findings and offered options, it is essential to review what you have done. Both before and after you present your report, answer:
- Would a non-expert understand your conclusion?
- What's the problem and did you resolve it?
- Did your report limit uncertainty for the decision-maker? Did you assist the decision-maker in taking action?
- Did you highlight parts of the questions you do not know and tell them how you are going to address any gaps?
- Have you defined time-sensitivity of the Intelligence and when you recommend revisiting the subject for an update?
- Can you summarize your answers in one or two sentences?
Always assume the process is flawed and try to grade it realistically. The intelligence itself is not flawed but the method by which it was gathered and analyzed may be.
Look at the Intelligence, define what you missed, and look for any gaps and errors. If there is a significant dispute in your findings, try setting up a "Red Team" tasked to find evidence which counters your answers.
SCIP: How often are insights in CI reports actually put into practice by the organization? For example, out of 10 recommendations should the CI team be happy if 3 are implemented?
GRAEME: An excellent question - and one I think you already know the answer to. On the face of it, three implemented reports is not an excellent record, but there is a unique opportunity to develop from there. However, there will be circumstances beyond your control. For example, your readers may not actually read the report (it sounds crazy but it happens) or value the hard work you have put into it.
It also depends on what impact those three reports have on the business. To find out if you wasted your time on those seven reports, answer these questions:
- Are the questions addressed in the reports really what the decision-maker needs? Were the insights useful?
- Did you spend enough time on the discovery phase? Did you have input? If not, why not?
- Were the reports forward-looking? Did they offer realistic options and solutions for the decision-maker?
- Did they answer why questions and the "so whats"?
- Why did the decision-maker reject seven reports? Do you know what they were looking for?
- Did you make the reports interesting enough for them to be read?
The insights in a report are the two pieces of bread in a sandwich. The best bit, the blackcurrant jam, are the options you offer decision-makers because of the insight. You must take the time to determine the possibilities. Imagine you are in the Oval Office briefing with the President of the United States. You explain what is happening in Syria. You go into the number of deaths, bombings, and refugee levels and the buildup of terrorists. It is accompanied by an excellent report, which you have spent all night writing. The President asks: "So what? What can we do about it?" A stunned silence will not go down well.
Decision-makers do not have time to sit down, read your report, and then think about the consequences. To increase your value and increase the chances of aspects of your report being implemented your reports must focus on the why's, how does it affect us and what should we be doing about it.
In other words, focus on the quality of the insight and the options you offer to the decision-maker. As an Intelligence professional, you are not the one making the decisions.
If your Intelligence is forward-looking, you have well thought out options; if the person with the responsibility to decide what to do fails to do anything, then it can't be your fault. But it is your problem, and a full review of the intelligence process is required. After all, a "review" is within the traditional intelligence cycle and perhaps the step most overlooked by intelligence professionals.
SCIP: Is a quarterly CI report suitable or is an annual one preferable?
GRAEME: When determining the suitability of a quarterly CI or annual report, you have to consider how much value your organization gives to CI. If you are just asked to write a yearly report, you have to ask if your organization values CI at all.
A lot can happen in a year, so leaving it to an annual report seems a bit pointless. Quarterly reporting is a little better, but you have to remember the power of CI is best used looking into what's going to happen, not what's already happened. When determining how frequently to report, consider the questions you've been asked to answer, the target audience, and timeliness of the content.
Most importantly, determine what questions are forward-looking, rather than backward-looking.
In the military, reports can be done almost instantaneously (what is happening with troop movements, what are they planning to do, why they are moving, who are they moving, what can we do about it,etc.) Some reports in the military consist
of just one line: Flash reporting decision-makers news on critical moves by the potential enemy. Imagine reporting like this for your company.
We recommend scanning for competitor and market activity and industry trends. Isolate new and emerging competitors by using software and primary Intelligence (talking to the right people and asking the right questions). For instance, define the competitor
- Key competitors - Your key competitors
- Emerging start-ups - The next big thing, those who have the potential to become critical competitors
- Watching - Potential competitors who have the resources to smash a hole in the market - For instance, Google developing something that blows you out of the water or Apple buying a key competitor
Intelligence reporting can be done when needed. Still, we recommend either weekly or monthly with the ability to create instant "Red-flag" reports when you need to reveal critical knowledge. In addition, give teams access to sales-based battle-cards
when they need them.
In summary, reports should be issued when you have something to say and when you have answered the questions. Not when someone up high says when they want it. If your management just wants an annual report and there is no value gained from
it, you best keep your resume in tip-top condition.
SCIP: What advice do you have on discovery for people new to CI?
GRAEME: The key to good Intelligence is asking great questions. So spend much of your time determining the key questions to ask. Here is a rough framework we use to develop questions to ensure we have a good shot of getting great answers for our clients:
- What do they need to know to solve the problem?
- What's the problem?
- What's the decision-maker trying to accomplish. What the goal?
- Who is the decision-maker?
- How can we limit uncertainty for the decision-maker?
- How can we create their needs into clear intelligence questions?
ASK THE RIGHT QUESTIONS
- Avoid yes/no questions
- Would a non-expert understand the question?
- Summarize the question in one sentence
- Make sure the question is inclusive - broad enough to cover what we think we'll need to cover, but narrow enough to provide a useful answer
- Do the questions meet the decision maker's needs?
- Is the question a need to know or nice to know? "Nice to knows" are always unnecessary time thieves.
- Will you be able to answer the question legally and ethically?
- Create a characteristics list that covers what you need to answer the question. characteristics (called drivers within Intelligence) could be leadership, recruits, people, money, weapons, Marketing, technology,
pricing, product benefits, sales teams, M&A activity
- Each question needs to have an associated characteristic. If you can do this, then is this the right question?
Taking all the planning into account, we then ensure a clear understanding of your needs, we have broken down your needs as follows:
- Situation - Describe the problem and what situation you are facing
- Mission - Our mission is to…
- Execution - We will answer the following intelligence questions to create valuable Intelligence, insight and recommendations (e.g. Intelligence question 1, 2, etc.)
Have a deadline and avoid wherever possible, the "can we just add this question to the project, please?" Say yes to one additional question, and pretty soon you will have 6 or 7 more questions to answer,
and in all likelihood, they will be a pain in the neck to answer and analyze.
SCIP: What are a few effective tactics field sales reps can use to gather competitor insights from prospects who hold their cards close to the vest?
GRAEME: The most important thing with sales reps is that they get the sale. It is how they earn a living. So, develop simple battle cards for them to read before they go into a sales meeting. Give them 3 or 4 killer points to destroy their competitor's value proposition, or make sure the sales rep knows of any bad news to drop into the conversation carefully.
Attend sales meetings with them or take a seat in your company's group sales meeting and listen. Ask how you can help them and then tell them how they can help you. Coach them on simple elicitation questions to get the prospect to open up and research the background of the prospect, so they get a rough understanding of them. Make a point of chatting to them over the water cooler.
Perhaps when the prospect picks you up from the reception mention that you saw a competitor's name in the visitors' book? Then listen to what they say. Develop it from there.
Another tactic is to ask your prospect a direct question about the competitor and then go silent. Either may not be the right things to do at that time, but you have to get a feel on the ground.
The last thing a salesperson wants to do is fill in more forms. They will have sufficient reporting already, so make sure any intelligence report is quickly done. Don't bombard them with lots of questions answer, just one or two. Reward them for finding the solution. I am sure an evening at a restaurant with their family on you will go down very well for finding a snippet of Intelligence.
It's all about wanting people to find information for you and liking you.
SCIP: How do you know or identify which questions to ask in order to go beyond the surface and uncover hidden insights?
GRAEME: The questions you ask to discover hidden insights are an essential part of a CI professional's armory. Take time to get them right, to ensure they are going to give you the answer. Have a good look to see if you can put some elicitation techniques into them.
However, in your planning, make sure you get a good understanding of the person who is going to answer the questions. Find out as much as possible about their background, their pastimes, their motivation to speak with you and, if you can, learn how they think and respond to specific elicitation techniques. But the most crucial thing when asking a well-structured question is to remain silent once they have given you a response. They will try and fill the silence with words.
If you are asking a well-prepared question via email, it is often tempting to send them a list for them to answer. I recommend one question at a time.
SCIP: When beginning with a hypothesis in mind how do you avoid falling into the trap of confirmation bias?
GRAEME: A great deal of focus is put on the collection of information. What war can teach you about business and Competitive Intelligence is that information collection is only part of the process. A bigger part is what you need to do with the information once you've got it: analyze it. How you use it to answer specific questions or how you make sense of that information to make it relevant to your customers (in a military setting, generals). So the CIA developed the Analysis of Competing Hypotheses (ACH) method to reduce the risk of cognitive bias within intelligence analysis.
ACH is a relatively simple analytic method based on the principle that it's a lot easier to prove a false hypothesis than to prove it's the truth. The truth sets a very high achievement bar. As in many cases, it could be impossible to answer a question but still go with the best argument you have. But give the decision-maker an indication of how certain you are. In reverse, though, it only takes a single piece of evidence to blow your theory out of the water. How much time do you save looking for the one piece of evidence that destroys your argument compared with the hundreds of hours it takes to prove your theory? First sort and rank the hypotheses. Then it becomes a process of elimination by removing options from your hypothesis until you're left with the ones that are not wrong and you have some supporting evidence. The effectiveness of ACH depends on what information you are using to rank systems. You need to take your time in developing this part of the ACH model. The first question to answer when ranking information is to ask simply, is it credible? Using the question "So what?" will also help here.
Also, when conducting ACH, take a leaf out of the detective book:
- Document everything
- Fix the timeline
- Follow every lead
- Everything is evidence
- Keep going
Remember each hypothesis can be broken down into the three-detective headline when trying to solve a crime:
- Is there any motive?
- Do they have the opportunity?
- And do they have the means?
If there's sufficient information to disprove any of them, you take the hypothesis out, and then hopefully you end up with one left standing. And that becomes the hypothesis which you have the best justification for taking as being true, essentially.
Finally, resist talking to the person who gave you the job in the first place. Keep them out of this process until you are sure you are on track. Industry experience, politics and their biases could cloud your decision.
SCIP: How do you focus on research and being an intelligence professional without being shackled by deadlines?
GRAEME: I feel your pain, and I do live in the real world, but being shackled by a deadline is ultimately your fault. Never agree to a deadline if you think you cannot achieve it.
You need to ensure that deadlines are essential for any intelligence project. A deadline is a deadline. Its like Rambo's helicopter leaving the roof at an exact time - if you're not on it, it's a long walk back to base, and his endless supply of ammo and rockets are very heavy. In the military, timely Intelligence is essential and could be a matter of life and death, so it should be the same with less complicated situations.
Deadlines tend to be a problem when the person who asked for the project adds a few "nice to knows" and some “CYJA'S - "can you just add's". With requests like this, try to get the deadline moved. Or suggest these new requirements may be completed under a new project, done after the current one. When doing research under a deadline, it is vital to take yourself away from the screen to allow time for thinking. I have whitewall moments and walks in the woods thinking about specific questions. It amazes me how many answers start developing when you take your head out of the laptop.
There are two ways to ensure that you meet deadlines. Firstly, give yourself an internal deadline for your research activities. It is common for professionals to research for way too long. Secondly, make sure the deadline you agree with your manager or client is realistic.
We often get product managers wanting a comprehensive Intelligence project with a seven-day timeframe. They have gone to 10 other CI companies for a quote too. It's a pointless exercise and we now respectfully decline. It's not worth the hassle, and the client rarely appreciates the effort it takes, especially if you are freelance and a supermassive company comes knocking with a stupidly tight deadline and a Holy Grail level of requirements. Only agree to it if you are:
- Sure you can achieve it
- Sure where the Holy Grail is
Tight deadlines usually come from a product manager leaving the research to the last minute, and their boss has asked for a progress report. They press the panic button and call you. If you are able to deliver, they will not appreciate it and think
you can cope with super tight deadlines. These projects are always the ones where you have to chase your money.
SCIP: What were some of the things you wish you knew when you began your intelligence career?
GRAEME: Well, to get into MI6 you have to go to the right school, university and be from the right family! But seriously, one of the critical aspects of an intelligence career (and any career) is to steer clear of the politics. Also, remember that most people are ordinary at what they do. Excelling means going that extra mile, doing a good job and making them see you actually care. And if it's not, don't complain, just get out and do your own thing and follow your passions.
Military Intelligence shows that you are nothing without the right team with the right skills. Do not have SAS Troops tasked to feed 500 people while giving the job of breaking an embassy siege to a group of regular soldiers just because they were around the corner and had just finished another project. The military has a saying "Prior Preparation and Planning Prevents "Particularly" Poor Performance." Real leadership ensures you have a great plan which doesn't include giving up at the first obstacle or changing direction because the boss told us not to. A career in the military or Intelligence teaches us to realize that mundane work is just as crucial as the sexy stuff. Also, it is what at least 90% of your competitors will not do as they cannot be bothered because they get bored with it.
Accuracy is a crucial demand within both Military and Competitive Intelligence, especially if there is a good chance that what you create could be used in the field. In the military, Intelligence has to be accurate, or it risks lives. The main thing is that Intelligence is about answering questions. Answers that your competitors would not like you know. Great solutions will only be achieved by spending time on developing quality questions.
Get away from your desk and talk to people in the business; meet the people who will know what you want to know. Get in the field and look at your competitors. See what is going on at their locations.
Finally, my motto - Keep It Simple. Within the military, everything is done as simply as possible. If it becomes complicated, then there is a significant chance that things will go wrong. The main reason for this is the communications chain. When under fire (real or verbal), we tend to get distracted. Remember, however, that planning for simple is rarely easy.
"Storm that hill from your right at 0600 to take out the machine gun post," is much better than, "We have conducted Big Data research and found that 70% of all field commanders would attack this hill from your
right and the optimum time to attack is 0600 because sleep patterns will be disrupted. Here is a SWOT analysis."