The Modern Graph Database

Buyer’s Guide
Delve into Functional Depth: Questions to ask Legacy
Graph DB vendors in your evaluation process

© 2023 ArangoDB, Inc.

 Introduction
In today's connected world, it's not just about the data;
it's also about how that data links up. Whether you're The Rise of Graph
creating a social network or recommending products,
those connections can be as valuable as the data itself.
That's why picking the right graph database is more
Technology
crucial than ever.
We’re all data-saturated. But just having a database
to stash and retrieve isolated bits of information is
falling short. Graph tech is revolutionizing the scene,
making it way easier for businesses and researchers to
understand and leverage the connections between
different data points.

This isn't just the realm of techies - it has real-world

applications. We're talking about better fraud
prevention, stronger cybersecurity, more effective
healthcare solutions, and smoother logistics, to name
just a few areas. In short, more accurate and better
insights into your data and business.

© 2023 ArangoDB, Inc.

 The Modern Graph Database Buyer’s Guide

The Many Use Cases of Graph

Databases
The applicability of graph databases extends across various sectors
and use-cases, some of which include:

1. Social Networks: Understanding and leveraging user connections for

better engagement and targeted advertising.

2. Recommendation Engines: Offering personalized suggestions based on

user behavior and item relationships.

3. Supply Chain Optimization: Mapping out the complex relationships

between suppliers, products, and logistics for better decision-making.

4. Fraud Detection: Unveiling hidden, complex patterns that indicate

fraudulent activities.

5. Semantic Search: Enhancing search algorithms to comprehend the context

and relationships between terms.

© 2023 ArangoDB, Inc.

 The Modern Graph Database Buyer’s Guide

The Cost of Getting it

Wrong
Choosing the wrong graph database can be costly and
detrimental. It may result in inferior performance, limiting
scalability and hampering your ability to adapt to evolving
data needs.

The stakes are high—inefficient data management can

lead to missed opportunities, and diminished
competitiveness.

An uninformed choice might lock your organization into

outdated technology, leading to increased operational
friction and potentially spiraling costs in attempts to
retrofit or migrate to a more suitable solution in the future.

The right decision is paramount.

© 2023 ArangoDB, Inc.

The Modern Graph Database Buyer’s Guide

Parsing Vendor Claims

ArangoDB is proud of the technology leadership and innovation it has
demonstrated over the years.

Customer after customer has shared stories about leveraging ArangoDB

to solve mission-critical problems after struggling with legacy graph
products.

In response to this leadership, legacy vendors have made claims that

their products can match the advanced capabilities of ArangoDB,
capabilities that global organizations have come to rely on when all else
has fallen short.

In the face of these claims, It is incumbent on prospects to ask the tough

questions that peel back the onion to help separate fact from fiction.

These questions are important to consider for another important reason:

ArangoDB will not always be the best fit given the unique requirements
of each project or application.

© 2023 ArangoDB, Inc.

 The Modern Graph Database Buyer’s Guide

Technical Evaluation Criteria

Simply switching from RDBMS or pure NoSQL to Graph
technology of any kind - while a step in the right
direction - can be fraught with risk and cost inefficiency.
The specific technology implementation each graph
vendor has undertaken can mean the difference between
project success or failure.

A careful and detailed evaluation of specific technical

capabilities is critical. The first step is understanding the
nuanced requirements of your applications and use cases
in FIVE key areas:

1. Data Management and Storage Model

2. Developer Productivity
3. Scalability, Performance, and Availability
4. Text Retrieval & Geospatial Capabilities
5. Integrated & Powerful Analytics & ML

© 2023 ArangoDB, Inc.

The Modern Graph Database Buyer’s Guide

Data Management & Storage

ArangoDB is proud to be the owner of one of the most powerful competitive
differentiators in database technology today - its multi-model capabilities.

There are countless technology teams that have gone on record to say that
this is the main reason they either selected ArangoDB in the first place, or
switched over when other solutions became too complex, too costly, or simply
lacked key functionality.

At first glance, the attempt to combine Graph, Document, Key/Value, and

Full-Text Search in a single database platform might appear overly-ambitious.
Yet ArangoDB’s technologists are proud of the fact that in addition to providing
this much needed flexibility, ArangoDB’s core graph functions are significantly
more advanced and rich compared to any other legacy graph vendor.

The simplicity of ArangoDB’s multi-model approach is undeniable. A single

query language to access everything. 100% data synchronization across all
models. Development simplicity. No performance or scalability hit as multiple
models are leveraged. No availability sacrifice.

© 2023 ArangoDB, Inc.

 Questions to Ask - Data Management

Legacy graph database vendors will claim to rival ArangoDB’s native, multi-model
capabilities. Here are some key questions to ask them to separate fact from fiction:

1. Your core database function is purely graph-oriented. In 5. Can we access these disparate database models (Graph,
order to match ArangoDB’s native multi-model capabilities, Document, Key/Value, Full-Text Search) simultaneously with a
what additional software layers are required to integrate full single query language and command? Or, are multiple, separate
document, full-text search, and key/value capabilities? queries and/or query syntaxes required? Any limitations on the
Anything above and beyond your basic graph database types of queries we can execute simultaneously across
instance? Any configuration changes needed to provide your disparate data silos?
“multi-model” capabilities?
6. Do we need to do any type of integration or perform any extra
2. You claim that your graph-only database can somehow mimic steps to combine results from these disparate queries executed
multi-model capabilities by attaching labels/attributes to against various data silos? How might this complicate our
graph nodes/vertices. How is this the same as providing the full development process? Would it steepen the learning curve?
capability and flexibility of JSON documents?
7. What kind of performance sacrifices should we expect as we
3. Are there any 3rd party tools or custom development required scale up our environment with these additional layers? What are
to simulate a true, native multi-model architecture? the added deployment considerations and complexities?

4. Do you offer fully-integrated full-text and geospatial 8. Is data kept 100% consistent at all times if we have related data
capabilities, or are separate products and integrations stored in different silos? If we want to attempt this with your
required? solution, what extra steps or development are required?

© 2023 ArangoDB, Inc.

The Modern Graph Database Buyer’s Guide

Developer Productivity
ArangoDB is centered around streamlining the developer experience,
standing out distinctly against the backdrop of legacy graph databases.

Constructed with usability at its core, it offers a clear and direct approach
to data management and querying. ArangoDB is in tune with the practical
needs and challenges of developers, delivering a workspace that is both
flexible and efficient.

There are many practical examples of how ArangoDB makes developers’

lives easier. Utilizing JSON data directly, avoiding complex data
transformations. Transparent scalability, so manual partitioning of data isn’t
needed. A unified query language, no need to learn multiple languages. A
fully-managed SaaS offering, no delays in getting started.

In this environment, the usual hurdles associated with database complexity

are diminished, allowing developers to allocate more of their time to
innovation and problem-solving. The result - significantly faster
time-to-market and richer, more functional app.

© 2023 ArangoDB, Inc.

 Questions to Ask - Developer Productivity

To match ArangoDB’s advantages in various functional areas, developers have to climb some steep
learning curves and spend extra time when using legacy products. Some questions to ask are:

1. How does your DB’s query language and tooling provide a 5. What steps are required if our developers want to enforce
unified experience for developers working with diverse data consistency and integrity, especially in distributed
data models? Can you show me? environments?

2. Please walk me through how your schema model 6. Kindly demonstrate how your database handles pruned
automatically adapts as data needs evolve and new use traversals, especially in terms of efficiency and resource
cases are added. utilization.

3. Can you walk me through the extra work required to 7. How can developers implement custom pruning
manually shard data? How often do I have to do this? conditions? Please demonstrate how developers can define
and modify conditions based on varying use case
4. How can developers optimize graph traversals in your requirements.
database, especially when dealing with high-density
vertices? Is there built-in support to ensure optimal 8. As it relates to executing concurrent algorithm processing
performance during complex traversals? on extremely large graphs, what special techniques are
additional steps must developers must take?
5. Can developers easily execute real-time analytical
operations without extensive configuration? 9. How do developers save time by relying on automatic
resource management, when managing extensive datasets?

© 2023 ArangoDB, Inc.

The Modern Graph Database Buyer’s Guide

Scalability, Performance, and Availability

ArangoDB stands out in its approach to scalability, performance, and availability
by offering distinct capabilities like seamless horizontal and transparent scaling and
data-model flexibility. ArangoDB has learned from the shortcomings of legacy
vendors who - while perhaps having a longer tenure in the market - are saddled with
legacy architectural baggage. These learnings have led to ArangoDB providing:

Transparent scalability, removing the need for developers to perform manual

sharding.

A unique set of smart optimizations, in particular for graph queries. This is

something crucial for going beyond POC-sized data sets.

A resilient fault-tolerance mechanism to ensure uninterrupted availability, allowing

developers to maintain a continuous workflow.

Adaptive query performance to ensure optimal efficiency for diverse data

interactions, freeing developers from manual optimizations.

© 2023 ArangoDB, Inc.

 Questions to Ask - Scalability, Availability, and Performance

Legacy graph vendors will cite their length of market presence and their partnerships and
integrations as “shiny objects” to distract prospective buyers from scalability shortcomings.

1. How do you achieve horizontal scalability, especially for 6. Please walk me through how your database automatically
large and growing datasets? How much should we budget fails over in case of an outage. If this isn’t possible, what
for additional infrastructure or operational overhead of any manual steps are required?
kind to try and mimic this with vertical scaling?
7. Please describe the process by which a developer has to
2. How does your solution optimize and maintain query do manual sharding of data if we want to create high
performance as the graph dataset grows? Is it linear? How availability clusters with low latency.
should we budget additional resources for this?
8. Please describe scenarios where I might lose transaction
3. How does your solution perform when handling mixed consistency in this setup, or when we are creating
workloads across different data models? What kind of replicas.
performance hit should we expect?
9. Can you please describe the process of automatic node
4. Please walk me through how your solution can dynamically replacement to ensure continuous service?
adjust its resource allocation as user demand and data
volumes grow and evolve. 10. Can I restore data from any given point in time?

5. How does your system ensure effective load balancing, and 11. How effectively does your database handle concurrent
what tools are provided for developers to monitor and transactions, and what provisions are in place for
manage load distribution, to maintain high performance developers to manage concurrent operations efficiently,
during high demand? ensuring data integrity and consistency?

© 2023 ArangoDB, Inc.

The Modern Graph Database Buyer’s Guide

Text Retrieval & Geospatial Capabilities

ArangoDB is at the forefront in pioneering the seamless integration of
geospatial and full-text search capabilities within a graph database.

For example, this groundbreaking initiative is transforming logistics, enabling

the integration of of real-time, precise location data to enhance supply chain
operations. In social networking, it's refining user interactions and connectivity
through advanced location-based services.

This approach is also reshaping retail, allowing businesses to create

consumer-centric solutions that leverage spatial data for personalized customer
experiences, and in healthcare, more accurate mapping of patient journeys.

ArangoDB cuts through the traditional complexities associated with

integrating full text and geographic data, paving the way for industries to
provide additional insight within applications that simply are impossible to build
with legacy graph products.

© 2023 ArangoDB, Inc.

 Questions to Ask - Text Retrieval & Geospatial Capabilities

Legacy graph vendors either completely lack these capabilities or have very rudimentary and
hard to implement feature sets that simply cannot meet the needs of text and geospatial apps.

1. How does your database handle the integration of spatial data 6. How efficiently can your database generate accurate,
natively to ensure optimal performance and minimal data real-time location-based recommendations, utilizing
redundancy while supporting diverse data models? spatial data and relationships to enhance user experience?

2. Do you offer more than rudimentary search capabilities, or must 7. Can you kindly describe how your product implements
I leverage an additional product like ElasticSearch for a robust robust geofencing capabilities to manage and monitor IoT
full-text search? How would I integrate this separate search data devices efficiently, ensuring precise location tracking and
in real time with my graph data? instant alerting?
3. How does your database support real-time tracking and 8. How does your database optimize spatial indexing to
updates of moving entities and provide seamless ensure rapid query performance and efficient data retrieval
synchronization between spatial and non-spatial data to maintain for complex spatial queries?
data integrity?
9. How can your database ensure flexibility in handling
4. How does your database facilitate in-depth spatial analytics, varied data models while maintaining the integrity and
allowing for the extraction of actionable insights from spatial accuracy of spatial data representations?
relationships and interactions within the data?
10. What options do I have to integrate with various
5. Please describe how your database conducts intricate spatial visualization tools to represent spatial data accurately
analyses to evaluate patterns, enabling the formulation of more and provide intuitive interfaces for spatial data interaction?
effective business strategies.

© 2023 ArangoDB, Inc.

The Modern Graph Database Buyer’s Guide

Integrated & Powerful GraphML

ArangoGraphML, integrated seamlessly into the ArangoGraph Insights Platform,
demolishes barriers between complex machine learning processes and users
with varying expertise levels.

ArangoDB’s intuitive approach to GraphML, combined with advanced algorithms

and rich libraries, enables effortless implementation of graph machine learning
solutions - in a fraction of the time and complexity of legacy graph solutions.

With ArangoGraphML, the focus is shifted from managing intricate infrastructure to

purely building and deploying models, ensuring optimal productivity and results.

Whether you are grappling with a surge in data, escalating your model complexity,
or extending your number of models, ArangoGraphML scales to meet your
evolving needs.

Most importantly, you can can meld full-text search, document, and geo-spatial
data into your ML modeling, resulting in far more advanced and detailed analytics
and insights extraction compared to competing solutions, all with minimal coding.

© 2023 ArangoDB, Inc.

 Questions to Ask - Integrated & Powerful GraphML

Legacy vendors might overstate the efficacy of their machine learning integrations, implying their
older platforms can match ArangoDB’s versatility and advanced handling of diverse data tasks.

1. How does your graph database allow users with limited machine 5. How does your solution simplify the process of preparing a
learning expertise to intuitively implement machine learning graph for machine learning, ensuring it's efficient and
solutions, while ensuring optimal focus on developing and requires minimal coding and input, while still optimizing for
deploying models, and not on underlying infrastructure and accuracy?
management?
6. What measures does your database take to facilitate an
2. Please show how your platform autonomously manages the end-to-end machine learning pipeline with high accuracy
training processes of machine learning models, allowing users to and minimal infrastructure management required for the
simply specify the data and training preferences and ensuring the users?
delivery of highly accurate results?
7. How does your platform support the use of a rich algorithm
3. How does your graph database platform handle spikes in data and library to enhance the machine learning capabilities within a
allow the expansion of model complexity without requiring graph structure, and how user-friendly are these integrations
modifications to the underlying infrastructure or management for both novice and expert users?
processes?
8. How does your graph database integrate machine learning
4. Can your graph database, through its managed service, and full-text search to enable in-depth analytics, allowing
dynamically adapt to evolving machine learning models and users to drill down into documents and extract meaningful
seamlessly oversee and sustain the underlying infrastructure? insights, all while utilizing advanced ML techniques?

© 2023 ArangoDB, Inc.

 The Modern Graph Database Buyer’s Guide

Conclusion
In conclusion, ArangoDB is reshaping the landscape of graph databases,
pushing the boundaries and setting new standards in technology and
application development.

It is important to note that ArangoDB may not always the best choice,
given the unique requirements of each project or application. Therefore, we
hope you can leverage this Buyer’s Guide as your evaluation handbook,
providing clear, concise insights and guiding you through the intricate
world of graph databases. Choosing the right graph db is not a “one size
fits all” proposition.

That said, we are confident - as our customers are - that ArangoDB stands
out, delivering straightforward and impactful solutions to diverse
challenges.

By leveraging the insights from this guide, we hope you are better
positioned to make savvy, well-informed decisions, unlocking the full
potential of whichever graph database you end up selecting.

We wish you the best in your Graph Database journey.

© 2023 ArangoDB, Inc.