In short
Financial research outsourcing is usually classified in two ways. One is by asset class or activity: equity research, credit research, fixed income, private credit, investment banking and M&A support. The other is by engagement model: dedicated analysts, project work, managed services. Both tell you what is covered and how it is staffed. Neither tells you whether you will be satisfied with what you get. A third classification does: value added.
Classified by value added, financial research outsourcing comes in two types. Lower value-added work delivers an output to a specification: a pitchbook, a set of comps, a model update. Higher value-added work carries an analytical opinion on the value or the risk of an asset or a company: a buy or sell recommendation on a bond, a credit opinion, an investment case. Lower value-added does not mean lesser. It is the right fit for a great deal of work and for many organisations. The higher value-added type is also available, and it is staffed and run differently.
- Product outsourcing delivers lower value-added work – defined outputs to a specification: pitchbooks, comps, model updates, monitoring reports. The opinion stays with your desk. It has served banks and asset managers for two decades and will go on doing so.
- Analyst outsourcing delivers higher value-added work – a dedicated India-based analyst, often a graduate of one of India's top few MBA schools, who works as an extension of your team and gives you an analytical opinion. You keep the decision and the sign-off. The analyst does the product work as well.
- An opinion is commissioned or it is not – there is some judgement in all work, but an opinion on value or risk is either part of what you ask for or it is not. That is why we count two types and no sliding scale.
- The measurable difference is turnover – average analyst tenure across the major providers is about 2.2 years. Frontline's is 6.6. A new person on the account every two years shapes the client's whole experience of the product model.
- Each is an equilibrium – a stable arrangement between provider and client. It rests on what the provider is built to deliver and on how open the client is, culturally, to a consultant's opinion.
- Neither suits everyone – it depends on the job, the organisation and factors only you can judge.
- The common problem is a mismatch – a desk that wanted an opinion and bought a product, usually because the second option never came into view.
What the two types are
The industry goes by several names: research outsourcing, offshore research, knowledge process outsourcing (KPO), investment research support. The names say where the work is done and who employs the people. They do not say which of two quite different things is being bought.
The line between them is whether the work carries an analytical opinion. There is some judgement in all work, a pitchbook included. But an opinion on the value of a company, or on the value or risk of an asset, is either part of what you commission or it is not. A pitchbook is lower value-added work in this sense. A buy or sell recommendation on a bond is higher value-added work. Many firms need only the first, and it is a sound thing to buy.
Product outsourcing is defined by the deliverable. The scope names the outputs, the format and the turnaround, and the provider is measured on whether they arrive on time and to specification. Typical work is pitchbooks and comps for an M&A team, model maintenance, spreading financials, covenant monitoring and periodic reports. It is low in judgement by design, though never zero. The desk supplies the thinking and the provider supplies the product.
Analyst outsourcing is defined by the person and their coverage. An offshore credit analyst takes a set of names; an offshore equity analyst takes a sector. The output still exists – notes, models, credit papers – but it follows from the analyst's view of the coverage, and the view is part of what you are buying. It is a junior's view. You ask what they think, you correct it, and the decision remains yours. Done well, the result is seamless integration with your research team.
Analysts produce products too. The difference is what defines the engagement, and therefore what the provider is organised to deliver.
Both are arrangements with a third party, so your outsourcing policy, your information-barrier procedures and your third line of defence will have a view on each. How a given arrangement is classified, and what oversight it needs, depends on the activity and the institution, and that is for your compliance function to decide.
A note on the words. Providers usually call these engagement models, and the vocabulary around them is used loosely. "Dedicated" means only that a person works for one client and no other, which is normal in both models. Pooling people across clients does happen in the market; Frontline does not do it, for compliance reasons. "Analyst" and "expert" are applied to almost everyone. The test of which model you are in is whether anyone asks the person for a view.
Dedicated or project is a different question
Most descriptions of the market divide it by staffing: a dedicated analyst who works only for you, or project work delivered against a brief. On that view you are buying either capacity or output, and the test is who directs the analyst's day. It is a useful distinction.
Ours cuts across it. A dedicated analyst can spend years on product work, to your templates and your priorities, without once being asked what they think. That is product outsourcing with a dedicated person. Analyst outsourcing begins when someone on the desk asks for the view and uses it. An opinion can be commissioned for a single project too; what matters is that one is asked for.
The two models side by side
| Product outsourcing | Analyst outsourcing | |
|---|---|---|
| Value added | Lower: an output to specification, such as a pitchbook or a model update, with no opinion attached | Higher: an analytical opinion on value or risk, such as a buy or sell recommendation on a bond |
| What you buy | A defined output, to specification | An analyst's coverage, including an opinion |
| What defines the scope | Deliverables, format, turnaround | The person, the book or sector, the working relationship |
| Analyst turnover | Average tenure about 2.2 years across the major providers: a new person on the account roughly every two years | Frontline average 6.6 years: the same analyst on your coverage year after year |
| Who is recruited | People hired for scoped tasks; graduates of India's leading MBA schools seldom choose it | Graduates of India's leading MBA schools (Frontline: the top 50 of roughly 1,300), with 2 to 12 years' experience |
| Who you talk to | A manager: a steady point of contact, but one who has not done the work. The people doing it change, often without meeting the client | The analyst, directly, with no manager in between |
| How they work | To the brief | Proactively: they bring ideas, argue for them and question yours |
| Where the opinion sits | With your desk | Drafted by the analyst, decided by your desk |
| Dedicated staff | Common, though some providers pool people across clients | Standard; it is not what separates the two models |
| How it is measured | On time, to specification, error rate | Whether the desk relies on the opinion |
| What it asks of you | A clear specification | Cultural openness to a consultant's opinion |
| What AI does to it | Lowers the price of the product | Speeds the groundwork; the opinion is still the analyst's |
| Suits | Volume, fixed formats, one-off projects, opinion kept in-house by choice | Continuing coverage, where context and a second opinion add value |
Who does the work
The two models recruit different people, because they offer different jobs.
India has roughly 1,300 MBA schools. Graduates of the leading ones seldom choose product outsourcing. Early in a career it offers scoped tasks and little contact with the client, since queries pass through a manager. That is not the growth or the challenge they are looking for.
The analyst model recruits from those schools; Frontline hires from the top 50. These are people with the trajectory of an elite young MBA anywhere. They are proactive, they come up with ideas and speak up for them, and they will question yours, as a good junior on a City desk would. The decision is still yours. Experience in the analyst model runs from early career, around two years, to twelve.
Turnover follows from this. Average analyst tenure across the major providers is about 2.2 years, so a client in the product model meets a new person on its account roughly every two years, and its experience is shaped by that. Its point of contact is usually steadier: a manager, who has not done the work. The people doing the work move on, often without the client having spoken to them. Frontline's average is 6.6 years.
The two figures are not a controlled comparison: one is an industry average and the other is one firm's. The useful step is to ask any provider for its own number.
Why each one is an equilibrium
An equilibrium is an arrangement that holds itself in place. Each of these does.
We say equilibrium where the industry says engagement model. They describe the same thing. An engagement model sounds like something a provider offers. An equilibrium takes both sides to hold, the client's as much as the provider's, and that is closer to what happens.
In product outsourcing the client sends scoped tasks, so the provider hires and trains for scoped tasks. The work arrives as specified, the client sees no reason to send anything broader, and the scope stays where it is. Every step is reasonable. The pattern dates from the first years of offshoring, when banks moved work cautiously, and it has been stable ever since.
In analyst outsourcing the client asks for a view, so the provider recruits and trains people who can form one and keeps them on the same coverage. The view improves with tenure, the desk relies on it more, the analyst has a job worth staying in, and tenure lengthens. The gap between 2.2 years and 6.6 comes from the work as much as from pay. We have written separately on what 2.2-year turnover costs the client.
The equilibrium sits between provider and client. It describes how the two work together and is no verdict on either. It is also why the arrangement is hard to change from one side alone. A client who starts asking a product team for views will not get good ones, and a provider who offers views to a client who wanted decks will not be thanked.
Who each one suits
Product outsourcing suits work where the output can be specified in advance and the thinking is deliberately kept in-house: high volumes, fixed formats, one-off projects, and any desk that wants its risky judgement made by its own people. That is a choice, and a sound one. If it gives you what you want, nothing about it needs changing.
Analyst outsourcing suits continuing coverage. A commercially aware analyst who has followed the same names for years notices what a specification would never have asked for. It is the higher of the two in judgement, and for many desks it is the better answer on satisfaction, output and cost. It is the older model in the City, where a desk has always asked its juniors what they think, and the newer one offshore.
Judgement can stay onshore if that suits you; it depends on the job and the organisation, and that is yours to judge. The higher equilibrium – analysts who form the view, you keep the decision – is also available, and we think more people should be at that level.
Why most buyers only ever see one
Most providers offer one of the two, so the other never comes into view. A buyer running a careful search compares product providers with each other on coverage, turnaround, licences and exit terms, chooses well, and gets what the request for proposal asked for.
Scale explains much of this. One of the largest providers states a workforce of more than 7,800 analysts, data specialists and technologists across 28 locations. An organisation of that size runs on standard processes and layers of management. That is the right design for delivering product reliably and in volume. Large providers have strong teams within them, but the structure is built for the product. An analyst who stays on one client's coverage for six years, and is asked for a view, is a different thing to organise.
So the useful question for a buyer is which model a provider is built for. Which provider is best comes second.
Where AI comes in
AI bears on the two models differently. Much product work can now be drafted by software and finished by a person, which lowers the price of the product. The product itself remains useful: desks still need decks, models and monitoring. What changes is the cost of producing them and the number of people it takes. The one job AI removes altogether is the pure checker: someone employed only to verify figures and sources, with nothing else to contribute.
In analyst outsourcing AI can speed the groundwork: gathering, cross-referencing, first drafts. The view, and the accountability for it, stay with a named person. As the price of product work falls, more desks are asking what they want people for. Some of them are asking about the analyst model for the first time.
When the model and the need do not match
Much of the disappointment with offshore research is a mismatch. The usual case is a desk that wanted someone who could see the bigger picture and bought a product. The provider was competent in its own terms and the onshore team was willing. The desk got what it asked for, and what it asked for was not the thing that would have solved its problem.
The reverse happens too: a desk that only needed decks, paying for analysts whose views it never reads.
| Bought product outsourcing | Bought analyst outsourcing | |
|---|---|---|
| Wanted an output to specification (lower value-added) | A match | A mismatch: opinions nobody reads |
| Wanted an analytical opinion (higher value-added) | A mismatch, and the common one | A match, where the firm is open to a consultant's opinion |
Where more was wanted, switching to another provider of the same model seldom helps, because the new provider is organised the same way. Moving from product outsourcing to analyst outsourcing is a change of regime. It means a different brief, a different kind of hire, direct contact with the analyst, and someone on the desk who reads a junior's view. The head of an onshore team with its own support centre in India faces the same choice, since a captive can sit in either model.
How culturally open are you to consultant views?
An equilibrium has two sides. The provider's side is what it is built to deliver. The client's side has less to do with what the organisation does than with what it is. Two banks doing the same credit work can sit in different equilibria.
The question that settles it is how culturally open the organisation is to a consultant's opinion. Some firms take an outside opinion readily and argue with it. Others prefer their opinions formed inside the firm and buy output from outside. Both are normal, and most firms know which they are.
In practice it shows in a few ways.
- Talking to the analyst. The model depends on regular conversation. This is much less of a question than it was 21 years ago, but it has not gone away. Some desks prefer to work in their own language and not in English, and that is a perfectly good way to run a desk. Analyst outsourcing will not suit it, because the exchange has to run both ways. Product outsourcing depends on it far less.
- Being questioned. An analyst will ask why, and will sometimes disagree. Some firms welcome that from a consultant and some would sooner not have it.
- Sharing the context. An opinion is only as good as what the analyst knows about what the desk is trying to do.
- A different kind of conversation, and no more time. In product outsourcing a process that one person used to hold is often broken into parts. Where the parts separate cleanly, that works well. Where they do not, the desk spends its time putting Humpty Dumpty back together again: where did that ratio come from, is the management biography complete, is a legal case missing. In analyst outsourcing the process stays whole, and the time goes on the conversation you would have with any graduate sitting next to you, whether they are 22 or 25. Everything that needs reviewing is still reviewed, and control and oversight stay onshore.
If the firm prefers its opinions formed in-house, product outsourcing is likely to be the better fit, and it is a sound choice.
How to tell which model you are in
What your scope document lists
If it lists deliverables, formats and turnaround times, you are buying a product. If it names people and coverage, you are buying an analyst.
Who you speak to
If queries go through a delivery manager, the provider is organised around the product. If you speak to the analyst most days, it is organised around the person.
What happens when someone leaves
In the product model a new person arrives and output continues after a dip. In the analyst model a departure is felt as the loss of a colleague, because the context went with them.
Whether anyone asks what the analyst thinks
If nobody on the desk has asked the offshore team for an opinion in the past quarter, you are in the product model, whatever the contract says.
What you would change
If the answer is "faster and cheaper", the product model suits you and the fix lies within it. If the answer is "I wish they understood what we are trying to do", you want the other model. This is what the higher equilibrium looks like in practice, and these are ten questions for evaluating an offshore analyst team.
Who checks the work in each model
In product outsourcing the control is the specification and a quality check against it, usually by the provider's own reviewers before delivery.
In analyst outsourcing the control is the one a desk already applies to its own juniors. On our side, senior London practitioners review everything a new analyst produces, then taper to sampling as the analyst proves reliable, with a senior always on call. On the client's side the analyst's draft goes through the same review, committee and sign-off as an onshore junior's would. Nothing reaches a client, a committee or a counterparty on the analyst's say-so, so adopting the model does not ask a manager to carry a new kind of risk.
Where the analyst model has limits
- One-off projects and pure volume. With no continuity of coverage, the product model is the better tool. We do that work too, on a different footing.
- Time. Good analysts are hired, calibrated to the desk and client-ready before they go near your work, and the value comes with tenure.
- What it asks of the desk. Someone has to read the view and say where it is wrong.
- Decisions do not move. Pricing, committee positioning, client relationships and final sign-off stay with the client in both models.
Whichever model you choose, write it into the scope, so that your governance, your review process and your provider are all describing the same arrangement.
Frequently asked questions
What are the two types of financial research outsourcing?
Product outsourcing is lower value-added: defined outputs to a specification, such as decks and models. Analyst outsourcing is higher value-added: an analyst who gives an analytical opinion on value or risk. In both, the decision stays with the client.
Is KPO the same as research outsourcing?
KPO (knowledge process outsourcing) is the older industry name for the same activity, also called offshore research or research support. The name does not tell you which model a provider is built for: product or analyst.
Which model is better?
Neither as a rule. Product outsourcing suits volume, fixed formats and judgement kept in-house. Analyst outsourcing suits continuing coverage where context matters. It depends on the job and the organisation.
Can judgement stay onshore?
Judgement can stay onshore if that suits you; it depends on the job and the organisation. The higher equilibrium – analysts who form the view, you keep the decision – is also available.
How do I move from product outsourcing to analyst outsourcing?
Change the brief first: scope by person and coverage, ask for direct contact with the analyst, and plan for someone on the desk to read and correct the view. Then check the provider is built for it.
What does AI change about research outsourcing?
AI lowers the cost of producing decks, models and monitoring, so the price of product work falls; the product is still needed. In the analyst model AI speeds the groundwork and the view stays with a named person.
I have tried several outsourcing providers and none worked. Should I give up?
Check the model first. If each provider delivered what the brief asked for and the desk still wanted more, the brief was for product outsourcing. Another provider of the same model will give the same result.
How do I improve my offshore research team?
Decide which model you want. Within product outsourcing, tighten the specification and the checks. If you want views and context, change the brief: scope by person and coverage, talk to the analyst directly, and ask what they think.
Is there a better way to outsource research?
There is a second way. Most providers sell product outsourcing: defined outputs to a brief. Analyst outsourcing gives you an analyst who drafts a view, with the decision kept by you. Which is better depends on the job and the organisation.
Part of: India-Based Analyst Teams – why they work
Compare: Offshore research providers compared
Go deeper: Offshore analysts can form a view. You keep oversight control. · What 2.2-year KPO turnover quietly costs you · How to evaluate an offshore analyst team