Logging interviews is a critical Lean Launchpad skill. Like the adage about the tree in the forest...if an interview isn't properly 'logged,' did it really ever happen?
Your goal in creating the log is the following:
- Create a sharable record of the interview to share with colleagues that were not present, instructors, your inventor, and advisors. Other people may be able to identify insight in a well written log post that you were not able to identify in the interview process.
- Reflect on the interview, capture key insights and next steps and tie each interview back to your broader themes and learnings.
- Capture what you thought at the time that will form part of your lessons learned narrative later on.
- Give yourself something to circle back to if and when you are short of insights in need of a pivot and looking for inspiration. (Hint: Insight often lives in the intersections between interviews. Beware of most recent data point bias...).
- Capture the status of the relationship with the interviewee. This can be critical if you want to circle back to someone later or end up moving forward with starting the business. Think off this as the CRM system for your business. A good CRM captures the details on the contact, but it also gives context and helps you pick up a conversation where you left off even if time has elapsed or different people are involved.
Below are some (unedited) examples of well logged interviews from previous classes. Notice the structure of the log and how 'insights' are both specific and actionable pieces of intelligence gleaned from the broader conversation.
Chief Medical Officer, Pathway Genomics
Joshua Maas Customer Interview 03/09/2017 Phone
KEY INSIGHTS
Principal Investigators mostly allow their biostatisticians to determine what analysis software is used in the clinical trial. He believes that the best use of T-Bio in clinical trial vetting would be in Stage II trials as a vetting measure for Stage III trials.
INTERVIEW DETAILS
Can you tell us your roles in clinical trial?
*Gave background - Involved in clinical trials at NIH and Cedar Sanai
There are two types of clinical trials.
1. Investigator – researcher comes up with question, comes up with protocol, submits to board and that's an interim process that goes back and forth for a bit and then assuming one has funding the investor submits study with IRB advertising, or other advertising techniques, patients come in. PI oversees the trial and is responsible to report any adverse events and submit a report of the trial. PI is the one of the main authors and will go out for publication and then will go out to society where PI goes and reports results and answers questions.
Once the funding has been granted, who are the decision makers on the types of software’s used?
- When a trial is initiated, the IRB looks at the anticipation of subjects likely to yield results, so they will ask you to do an analysis on the number of patients needed.
- Generally most investigators will consult with bio-statistician to define this
Is a bio-statistician someone who is hired at the beginning only or only throughout the trial?
- depends on the type of trial
- for lab trials they will hire an outside bio-statistician to come in
- usually the statistician will come in and stay throughout the entire trial
Is bio-statistician involved in screening process?
- recruit a patient to the trial, explanation of trial, inclusion, exclusion of trial, go through formal consent process, then you make secondary decision whether they are useful or not for trial
- there are software’s that can screen patients through survey questions
Do you ever use genomics to screen?
- This is getting in to a very interesting area.
- If they know a drug will only work on about 25% of population then, yes, they would use genomic testing and sequencing.
- Can’t get to genomic info unless a patient is given permission to do that
Do you have to find patients, or do they find you?
- in our conformed since, New York is opt in, everywhere else is opt in
Do you know the cost of screening patients?
- No, I honestly don't
- I can tell you if we are recruiting patients through a doctors office, we are talking about $250-$300 for patient for private doctors
- If you go to a health fair, and you recruit then and there, that's a very cheap process. If you can pick up 50 patients you’re looking at a very cheap process
- if you use a CRO to find new patients and run the study, that's a very expensive process.
If you use a CRO do you have final say in screening?
- absolutely this is an extremely extensive process
What type of analysis software are you currently using?
- most of what we are doing is coral to studies and observational studies
- software we use is desktop statistical packages
- we will do RRC analysis for specificity
- we are involved in studies looking at pharmacogenomics looking at anti-depressants with a lot of data analysis so we are partnering with statistician who is very versed in this area though I don't know what exactly he is using
[Explanation of Pine]
Is this something that we would want to talk to bio-statistician about or do you have insight?
- simplest would be if you do genomic testing for BRCH1 genes and others associated with breast cancer in women or men, and you find hits, those are the patients you may want to enter into prevention trials.
- there is an example where you would use gene screening for picking patients at high risk for getting a disease and putting them in a preventative trial research
- you can use genes for screening for some trials but you have to figure out what the likelihood is there will be enough patients and is their disease different from the patients at large?
- Place to do discovery work in in phase 2 clinical trials.
- good to incorporate running genetic testing in phase 3 clinical trials
- FDA requires after phase 3 that only patients with those specific genomes gets that drug
Is there a market for a platform to be used in phase 3 genomic sets?
- its competitive but if they do it well and have genomics and proteomics availability then yes I think they can approach drug companies for this
- it cost $50 to run test, we charge you $75, all data is yours
- we will collaborate with you on this trial and do it free but want a certain percentage of the backend profit from the trial
- revenue sharing
Anyone else we could talk to?
- there is a guy Eric Brass, at Harvard Medicine – MD Pharmacologist that helps FDA approve trials
- niche companies that help companies get ready for FDA presentations – head of company, Sidney Dabossi, 3D Communications in Washington, DC
Must Have
R$ | Usage Fee
Nice to Have
KR | Access to customer Data
CS | FAKE VARIABLE (to update contact information)
Don't Care
Director of PV Product Development, Spectrolab
Elizabeth Spencer Customer Interview 03/07/2017 Phone
KEY INSIGHTS
A simplified- and visual- explanation of how light travels and is converted to energy/heat from beginning to end of the receiver, coupled with points addressing how it’s more efficient than traditional PV, could make a compelling addition to an MVP.
INTERVIEW DETAILS
What we thought before this interview:
Jim is involved in solar cell development assistance with the rest of the team- Spectrolab specializes in cell development. They are one part of the value chain- the cells are then incorporated into some sort of CPV utility system.
What we learned from this interview:
This was a great interview to learn some of the high level science going into this project. Jim did a great job of breaking it down for us in simpler terms (and some not so simple!).
Traditionally- PV captures light at a certain ‘band’ within the overall spectrum. What isn’t captured and converted into energy, is dissipated (in some way) as heat/energy. The novel thing about this tech is that it utilizes transmissive (layered or multi-junction) pv cells- light is captured efficiently from ‘highest’ to ‘lowest’ spectrums by ordering PV cells that absorb those various spectrums best, in order. Sort of like having your cake and eating it too. At the end of the chain of PV cells the leftover (uncaptured) infrared light feeds into the thermal receiver- generating heat.
Jim told us in the early stages of the project, the primary consideration was dispatchability- making sure you match the demand for power on the grid with the amount of power available. The initial idea was to use the heat generated as a ‘thermal battery’ that could be dispatched (and converted to electricity) during peak hours, when the sun isn’t shining, etc. However, the project pivoted towards finding a direct application for using the heat.
Jim also brought up the low cost of traditional PV as one of the challenges for CPV. Ten years ago, silicon panels were about $1.70/watt. Today, silicon panels out of China are $0.40-0.45/watt. Many companies investing in CPV were able to compete with traditional PV at $1.70, but not at $0.40.
What we are doing next as a result of this interview:
Jim brought a point that has been a recurring theme in a couple other interviews (Bob O’Hara and Paul Breslow)- The cost of traditional flat PV on a single axis tracker has gone down dramatically in the past decade, coupled with an increased efficiency in the modules, and it’s also the incumbent technology. The challenge for CPV is that it must differentiate itself and find applications where it is demonstrably better than traditional PV.
Our basic value proposition is that our product provides a CHP solution- but we may need to rethink/pick-apart our value props to make sure that we’re aligning them with differentiating our product from traditional PV.
Must Have
Nice to Have
VP | Combined heat and power