blogCustomer Spotlight: Mizkan Holdings

Mizkan Group (Mizkan for short) is a company with a long history of manufacturing and selling food products that stretches all the way back to vinegar production during Japan’s Edo period. Today they strive to better support consumers’ daily lives while maintaining a deep respect for local food culture around the world. In 2019 Mizkan launched a new brand "ZENB" based on the concept of contributing to the health of people, society, and the earth. They’ve received a lot of attention as a long-established company with a 210 year history now embarking on a new project looking towards the future of food.

For this edition of customer spotlight, we spoke with Mr. Manabu Konishi who is part of the new business development team responsible for the ZENB project. He shared how he used competitive intelligence from Amplified as a patent liaison to support the research and development behind ZENB.

Mizkan was founded in 1804 to manufacture and sell food seasonings starting with vinegar as well as food products such as natto. We’ve been expanding abroad aggressively with our overseas revenue now making up more than half of Mizkan Group's total sales. For example, we manufacture and sell Bertolli brand pasta sauce in North America as well as Branston sweet pickles and Sarson’s malt vinegar in the UK.

In March of 2019 we launched a new brand called "ZENB". ZENB is a new concept food brand that seeks to better the health of people, society, and the Earth not just now but ten years from now. We do this by bringing out the natural flavors and nutrition of the ingredients without relying on animal-based flavorings or additives and by using the entire plant including things that are usually discarded like skin, core, and pods to reduce food loss and environmental burden.

I’m in charge of intellectual property for the project but I sit closer to the R&D side where I act as a "patent liaison" to bridge the R&D and IP teams. Specifically I have two main roles.

The first role is patent filing related work and securing our IP rights. I’m working to architect a network of patents by holding discussions with all researchers every three months. In these meetings we seek to identify patentable inventions and budding ideas that are likely to become patents. The second role is about using competitor’s patents as a source of business intelligence to help our researchers come up with new ideas and overcome technical hurdles in the R&D process.

The strength of the R&D-oriented patent liaison role is that we get involved with the researchers and engineers from the very early stages of R&D when the initial concept is decided. If we find prior art after this phase progresses it’s a huge problem, so I can support them by listening to the developer first, digest what they are saying, and then sharing some practical relevant information early on. Patents don’t always have the final “answer” but at a minimum incorporating prior art at a very early stage leads to faster development speed. Also when working alone, engineers sometimes get stuck. I’ve noticed that in these situations just talking with me to explain the problem they are trying to solve often helps them organize their thinking and get unstuck. I also focus my questions purely on the problem without considering the annoyance or feasibility of potential solutions. So that also helps us to illuminate the key problem more clearly.

I didn’t have any IP background originally but working in various departments such as R&D, product planning, and marketing over the past 12 years has helped me cultivate information gathering techniques, grasp the company’s overall strategy, and understand the current situation. All of this is useful for brainstorming with developers so my past experience outside of IP has become my strength in IP work.

Actually no, it’s something that I proposed. The reason was simply that I kept wondering, "Why doesn't everyone read the patent first?". I began reading patent information after I was put in charge of this project. After actually reading some patents, I thought, "Wow, patents are such a treasure trove of information!" Just in the first 3-4 months of this we found several times that our current research work was already written in competitors' patents from 5 or 10 years ago! As a result of that experience, I decided that we needed to make the use of competitive intelligence more systematic and that has become a core theme for my work.

I find competitors' patent information is really useful, especially for new businesses. In product categories like sauces where we’ve been working for a long time, Mizkan has built up a lot of internal experience so instead of trying to get something from competitors information there’s a strong trend towards trying to get more out of our abundant internal knowledge. 

However, especially as we’re getting into more and more completely new categories faster than ever, utilizing competitor information becomes essential. For example, some recipes and manufacturing methods can be very useful in early stages and if our concept isn’t new it’s easy to quickly find that out. In the middle stages of development I find that competitive intelligence is useful for improving products to handle production defects and meet customer requests. For example, if a customer wants a better texture and mouthfeel, you can find information about others who had the same problem and learn from that.

On the other hand, I think our development ability is really amazing in areas where we have deep know-how like flavor making. For example, a person who had been in charge of developing hot pot soup bases joined us and made an incredibly delicious soup noodle seasoning for ZENB’s soup noodle line. I was shocked at the depth of flavor even though it doesn't use any animal-based ingredients at all. The product concept is catchy and attracts the market’s attention but I feel the real strength of our product comes from our the know-how and fundamental technologies to create a good taste that we've developed over our long history.

After joining the project in 2019, I learned that there are some AI patent search tools. Although patents contain a wealth of information that is useful for research and development, it’s not easy for people with limited IP experience like me to extract the right information. I always felt extremely frustrated by that and then I found out there are AI tools that are easy to use and don’t require search strings. After researching and comparing several AI tools, Amplified was the one that matched my needs. Also the pricing and purchasing is simple and flexible so I didn’t have any reservations about getting started right away.

Broadly there were three reasons I was attracted to Amplified’s solution:

1. The first is the UI. It doesn’t require search strings and the UI is intuitive and natural. Although there are other tools that use AI to produce search results, I felt a big difference because Amplified’s design allows you to actually read the search results directly in the interface and then leverage those results to search more.

2. The second reason is the search accuracy. Amplified doesn't limit the result set, right? That’s important. Even if it’s easy to make a search, the attractiveness as a solution is slashed in half if the search accuracy itself is low. But the similar documents that I found confirmed Amplified’s results are highly accurate which is especially impressive given the result set is not limited at all.

3. And the third reason is the sheer processing speed. There are quite a few cases where the sentences and keywords used in the first search are not very good. Especially when you are not yet familiar with the tool and don’t have an intuition for it yet, I think it can be difficult to get good results on the first try. The other tools that I tried often took tens of minutes up to even several hours to load search results but Amplified loaded results immediately which means I could observe when my search approach wasn’t working well or when the prior art didn’t include information in this way. Fast loading lets me quickly adjust my search and then check results again so speed was a big reason for choosing Amplified.

After introducing Amplified to our work, I feel that the psychological and practical hurdles against patent searching have dropped. For example, when an unfamiliar technical field comes up in a meeting I can use Amplified in downtime immediately afterwards to rapidly investigate the same way I use Google to quickly learn about something unfamiliar. It used to take 30 minutes to 1 hour to create a proper search query before reading results, but with Amplified now it only takes about 15 minutes to get an overview of relevant patents.

My focus now is on expanding direct usage of Amplified to more people throughout the company. So far I’ve done the first search based on interviewing the researchers but I think the ideal process is for the researchers and engineers to start the search on their own.

To make that happen, R&D members need to experience patent information helping their work first hand. Then they will naturally want to search on their own to solve problems. People won’t go to the trouble of reading patents or trying new tools if they don’t see a benefit.

Amplified’s recent addition of Viewer and Annotation features are helping me to overcome this obstacle though. I can add Viewers for free and share project results with them in Amplified. Then we can highlight and comment on specific sections of the patent to easily highlight and share information. Using these features I’m now trying a new workflow where I share my search results as-is to the engineer directly in Amplified and then ask them to review in the tool.

If researchers and engineers start to benefit from patent information and build up success stories then I think they’ll start trying to research proactively and I’ll also benefit from seeing new perspectives through the comments they leave. With that said, it’s hard to solve problems with the information just by passing along search results so what I am doing now is kind of “gluing” together the information in patents and the challenges they are facing. When sharing the search results I also ask things like “Can't we use this idea to do something like this?” or prompt some brainstorming discussion based on different ideas picked up from Amplified results.

The first step is just continuously fostering awareness of how to use patent information to solve R&D problems. As we accumulate more and more success stories, R&D team members will increasingly realize “oh, I can use this information in this way to solve problems – and I can do it on my own.” Ultimately when everyone is using Amplified that will create a unique patent database just for Mizkan that is really interesting. We can create our own custom knowledge base that adds the ideas and perspectives of our company on top of the patent information. For that purpose, I think it will be necessary to develop some internal guidelines for writing comments that will be useful in the future. But even without that just leaving a comment shows you that maybe someone else in the company saw the same patent a year ago and I can see what they were interested in at the time. I think this kind of information sharing can be a great tool to facilitate and promote communication within our team.

Patents don’t have explicit answers to R&D questions so I think it would be good to have a way, a mechanism, for drawing awareness to some key hints in the patent information that help to come up with new ideas. For example, if you read the search results and find a sentence or comment that interests you, then perhaps clicking on that to immediately see more patents related to that particular thing. Also since the comments we leave accumulate over time, I’d love to have a function for searching through the comments themselves.

There are quite a lot of tools to search for patents but there aren't many that use search results to come up with ideas. I think it would be interesting to add features not just for reading the patent but also to facilitate applying the information after reading.